I’m going to hazard a guess that it is because it undermines the entire reason the Sonos system exists? Essentially, that delay allows the entire “whole home system” that Sonos has based their business on to play in sync across all speakers. Rewriting the base software/firmware of their entire system might be challenging, in order to deal with a minority of users. 

Is it not ideal? Sure. But you have to consider where Sonos wants their bread and butter to be. Is it a networked home music system, or do they want to only be a home theater system company? I would guess they’re following the money.

There’s ups and downs both ways, but i don’t see Sonos making a major shift from where they built their company. 

It does, however, allow you the opportunity to build your own specialist company, that provides this particular feature without needing to include the music aspect of streaming across multiple devices, but only the home theater aspect. Perhaps you could be wildly successful.

You would need to increase the delay for all SONOS Rooms except the PORT driving the SONY. RAM was much more expensive when the SONOS protocol was introduced in 2005 and it is not practical to upgrade the RAM or processor. If current units implemented this sort of feature, they would not be as compatible with older units. SONOS is also attempting to keep setup as simple as possible. Variable delays in each unit would be an unreasonable complication for most users.

If you like complicated, you can stretch out Line-In’s latency. If you introduce another PORT to be the source of all music on SONOS, then connect it’s Line-Out to another Line-In, you have some ability to increase the latency on that second PORT’s Line-In. Likely you would then need to increase the latency on the SONY to align with the new SONOS latency. Operationally, this would be somewhat contorted.

I have no idea about Sonos’s business model.
But I don’t see Sonos offering a proper video compatible surround sound sound system.  So if their main focus is basically about selling hardware, like Apple, then they should accept that they will have to interface with other company’s systems and provide compatibility features; such as AV amp compensation delays. 

Sorry but I think you have missed the point.
I stated that In my neighbour’s house most of the rooms are acoustically isolated. The only problem is between the acoustically connected lounge and kitchen/dinner. Therefore to fix the problem you only need to increase the delay to the kitchen/diner Amp. 
And yes, it is probably impractical to add RAM to existing hardware - unless it is socketed. But newer hardware could have more RAM. Given the price of Sonos kit and RAM the RAM cost is insignificant. 
Your argument that newer units would be incompatible with older units is simply wrong. Adding the feature would simply mean that newer units have the feature and older ones don’t. They would still co-exist on the same network. The older units simply wouldn’t offer the feature. 
If Sonos’s business model is to sell more hardware then providing features that only work on new hardware sounds like an excellent plan. I believe that is why my neighbour has a whole stack of older Sonos kit that he now needs to sell because it is not compatible with the recently acquired stuff. I’m not a Sonos expert, I’m just reporting what I see. 

And as for keeping setup as simple as possible, yes let’s do it for simple systems. But when you have more complex systems then you need an ‘Advanced’ menu. That is why Sonos’s default operating mode is by WiFi. You don’t need to run wires anywhere. And why Sonos sells the Sonos Boost - because sometimes you need a dedicated WiFi network. 
Actually my neighbour’s system is so complex because of the number of systems, walls and other WiFi devices that to get it to run reliably we have had to hard wire it using the Ethernet connections on the back of each unit. Umm, if setup has to be simple then why does Sonos provide hardwired Ethernet ports?

Ahh yes, using the Line Out and Line In connections. These are Analogue signals. The whole point of Sonos, and other modern music delivery systems such as Tidal, Qobuz, Spotify, ROON and Meridian is to keep the signal in the digital domain for as long as possible. Thus avoiding distortion and noise (yes I do know about jitter etc). Feeding a digital signal such as from Tidal into a Sonos device, doing a D/A conversion, feeding that out of an analogue port into a different analogue input, doing an A/D conversion, processing that signal, doing another D/A conversion and then feeding it out to speakers is just perverse. 
And all for the sake of not providing a simple delay function!

My own home’s multi-room HiFi system is based on ROON. I’ve just checked and ROON offers a variable delay of up to 1000ms specific to each individual Endpoint. 

Similar comment could be made with respect to the SONY A/V. Why does it need to add a delay?

The delay is caused by the software processing needed to generate the 5 or 7 surround sound signals from the original stereo. The video signal then has to be delayed to match the sound to get lip synch. 
If you change the processing mode to say Direct then you take out the surround sound software processing and hence the delay changes. To maintain lip sync the amp either takes the delay out of the video path or adds delay back into the audio. Reading the other threads on the topic it seems the more common option is to add audio delay. As the data rate is lower for audio than video I guess that adding audio delay is easier. 

You’re quite right that this issue has been doing the rounds for years. Technically it might be possible to come up with a scheme that could satisfy the small proportion of folks with Connect/Port + AVR and another room grouped audibly nearby.

Why have Sonos not offered a solution? Simple: priorities. There’s a never-ending list of development requests, and it evidently makes zero commercial sense for Sonos to divert precious engineering resource to address the issue, compared to the return on efforts directed elsewhere. 

The Port apparently sells in relatively small quantities compared to the speakers/soundbars. Some probably go to installers feeding third party home-integrated systems, some to HiFi buffs wanting to Sonos-ise their existing mid/higher-end gear, some to those with an AVR. And of that fraction of a fraction of total Sonos sales, a fraction may sometimes want to use Port + AVR with an audible room grouped nearby. It sure sounds like a tough sell for a product manager, compared to all the competing demands. 

If the source is stereo, use the direct/bypass mode on the Sony Amp, to avoid the processing delay on the Sony A/V Amp? 

According to the OP, direct mode makes no difference.  I find that strange. Perhaps to all the fractions correctly identified by @ratty we should add 'the fraction of A/V amps for which the problem cannot be solved by flipping a switch'?

There is something about the above that does not seem right, or I am missing something...

Are you suggesting that where using the stereo mode on an AVR to get it to sync with Sonos - and most  AVR kit usually does that well if said mode is there as a feature - lip sync is thrown out of whack in a discernible way? 

I have not heard of that con till now as the price to be paid for perfect audio sync. I suspect that for lip sync to be affected, more milliseconds are needed than those needed to be eliminated to get rid of the echo.

I read it as direct mode does resolve the issue, but then neighbor can’t have the A/V receiver process the audio to create artificial surround sound channels, as desired.  Also seems as though they don’t want to toggle direct mode on and off when switching between  TV audio and audio coming from the Port. 

That’s all understandable, but I don’t really see that as a problem that Sonos needs to address.    While it does appear to only effect a fraction of Sonos customers, those tiny few who would take advantage of it are not any more likely to buy Sonos hardware for it.  I mean, those who have this issue aren’t going to buy Sonos HT equipment, because Sonos would have just made it easier not to, and because the desired feature of simulated surround sound isn’t something Sonos offers.  Would a customer decide not to go with Sonos for wireless home audio for this reason?  Doubtful.

So I would think Sonos would spend more dev effort increasing their functionality in home theatre, rather than easier to use someone elses products.  However, I don’t see them ever doing simulated surround sound from stereo input.

Ok, I’ve been back and worked through all the settings again. Setting the Sonos input on the Sony Amp to Direct mode does reduce the delay to an acceptable level. But you cannot use the surround sound features of the amplifier without re-introducing the delay.

But the source is stereo? Surely you want to reproduce the original sound, and don’t want to mess with it to create a ‘synthetic’ surround sound?

The delay is introduced by the AVR to deliver these effects, and I don’t think the delay is large enough to create lip sync issues by toggling the stereo button.

To the quoted part, I suspect that even with a stereo source, and using the AVR dsp, the centre channel would still work, and if so, deliver the great benefit of controlling dialogue levels precisely when needed. Of course, the cost of that would be losing sync with Sonos.

One has to choose between these effects and perfect Sonos sync. I don’t see that ever changing.

@craigski You state “But you cannot use the surround sound features of the amplifier without re-introducing the delay.” On my old Marantz (if I remember it correctly) the “Direct” setting worked per source. So you would only lose the surround features on the Sonos (that is stereo anyway). 

I didn’t state that, it was OP 

This must have been a convenience feature to avoid having to select direct where the source selected was the Sonos device wired to the AVR. You would, I expect, still get only the two front L/R speakers wired to the Marantz firing in this mode, but in perfect sync with any Sonos speaker in the same group as the Connect/Port wired to the Marantz.

The source of audio to the AVR in this case is the Port, not TV.  There are no lips to sync with.

I suppose their could be dialogue, such as podcast or what not streaming through the Port and Sonos network in general, but I really don’t think there would be a need to activate the center channel for these stereo sources.

A surround sound processor operating in a mode such as Meridian’s Trifield put common L & R sounds through the centre speaker, leaving L or R specific sounds on their respective channels. Thus anything centre stage, usually a singer or speaker comes out only on the centre speaker. The effect of this is that vocals are much cleaner and it reduces the criticality of the stereo ‘sweet spot’. Both people sat on the sofa get good vocals! Basically it reduces the audio clutter produced by sending the same sounds to independent L & R speakers. Its why I prefer stereo music played through a surround sound system.

I get the idea, it’s just seems that a center channel for clear dialogue seems less important for stereo music than for TV watching.

Indeed, and a full on duh moment for me. I was lulled down a wrong lane by my use case where lip sync matters - I have a Connect receiving TV audio via line in, feeding a stereo amp connected to two quality speakers and a sub in my patio. And for parties, the sound can be piped around the house to other Sonos units as well, in perfect sync including lip-sync with both TV video and streamed or NAS music from Connect as a source. All that is needed for this to work flawlessly is for all Sonos units to be ethernet wired to the core network to allow stable grouped play with delays set to the minimum, a requirement for lip sync when the TV audio is a feed.

For a plain vanilla set up of Connect to AVR where the former is a source of music, lip sync, as quoted, is indeed irrelevant.

But the OP makes an interesting point about the centre channel still being useful in such cases; I have never tried out that to comment on whether that brings about an improvement in the music listening experience because the last time I used an AVR was a couple of decades ago. I realised then that a full wired HT set up was more clutter than what the added audio effects were worth; my TV sound improved enough for my needs by just a stereo amp and two quality front speakers, and that has remained my TV approach since then. If I want to watch Top Gun/Maverick in all its glory - which is a rare need - I go to the multiplex nearby.

FYI: https://en.wikipedia.org/wiki/Trifield

Interesting, and this was my finding when I was seeing if I can usefully deploy a spare play 1 unit a year ago. In my Connect Amp + Quad 11L stereo listening zone, I placed the 1 unit at the centre between the Quad speaker pair, and grouped it with the Connect Amp. It took me some time to get both at the same sound levels, but when I then sat back and did some critical listening, all I found was that the music was louder - no change perceived in sound stage or music quality. Since the Connect Amp has enough headroom on its volume control in that space, louder was no benefit, and I put the 1 unit in a drawer where it still languishes.

I don’t think that any fancier tech than what this little experiment contained will bring about noticeable change to the listening experience in my space.

If I understand correctly, this setup with the extra Play:1 means that the Connect Amp will be correctly splitting out Left and Right stereo signal to the Quad speakers. But the Play:1 will be playing both the left and right signals combined into a single Mono channel. So placing the Play:1 between the two Quad speakers will just mess up the stereo image as you will now have the Left only sounds coming from both the Left speaker and the central Play:1. Right only sounds will come from both the right and centre. Depending on volume levels, I guess this would sound like a wide mono stage? Certainly not an improved stereo stage.

The problem with a 2 channel ‘stereo’ setup is that centre stage sounds come from both left and right speakers. Hence if you are not sat exactly equidistant from both speakers then the sound stage is distorted.

A Trifield or similar effect produced by an AV amp, results in sounds from the left of the stage coming only from the left speaker, sounds from the centre only coming from the centre speaker and right stage sounds only coming from the right speaker. The result is that the centre sound is ‘cleaner’ and the sound stage is less listening position dependant.

Its not quite as simple as that as the centre speaker is usually a small unit that has poor bass response. So the AV amp leaves bass on the left and right channels. Or sends it to a separate subwoofer; which allows smaller, less obtrusive left and right speakers as well (often preferred by ‘the other half’ sharing the sofa:-))

But ….. the software signal processing needed to produce this better ‘stereo’ sound stage results in the sound from the AV amp being delayed relative to acoustically close Sonos only systems. Which is why we need a function in the Sonos setup to provide a variable delay to compensate for the addition signal processing in the AV amp.