Audiophile Sonos Connect?

I suggest you read the oft-referenced article from Monty Montgomery of Xiph (the developers of, amongst many projects, FLAC).

Sonos have stated a couple of years ago that, whilst they can see a slight case for 24-bit (and the current hardware could apparently support it in theory), higher sampling rates make no sense.
https://www.whathifi.com/features/sonos-plans-brighter-and-brighter-wireless-music-future : "the math just isn’t there"

Hey there, todd1010. Thanks for posting your feedback. I am happy to forward this along to the team for consideration. Keep it coming.

to connect to my two channel listening room to handle hi-res music playback.

The worst possible reason for replacing the Connect or for Sonos to make one; hi res is a red herring. I can understand wanting a new Connect for more features and/or lower price, although the latter is very unlikely given the effect it will have on play unit sales.
But anyone that must drink the Hi Res Kool Aid can do it for just USD 35; get a Chromecast Audio.

Pedantically (and elaborating, not correcting anyone), Monty didn't create the FLAC codec, Josh Coalson did. (Developer can mean many things, of course) Monty did create Ogg Vorbis and many of the other Xiph.Org projects. At FLAC version 1.2 or 1.3, it was transferred to Xiph.Org, and had been in essentially maintenance mode for a few years before the transfer. Josh did decide to use the Vorbis tags as the FLAC tags, so there's obviously been interaction between Josh and Monty. Open Source audio codecs was a small world in the early 2000's.

Also, there has been "updates" to the Connect, at the same price point and explicit features. [https://wikidevi.com/wiki/Sonos_Connect_(S15)] We don't know if the internal RAM is the same as previous iterations or not, and internal RAM has been one of the major differences from the earliest Sonos units (ZP80 had 32MB, Play:1 has 256MB) other than the wifi capabilities, ZP80 had 802.11b/g 1x1, Connect has 802.11b/g/n 3x3. Internal RAM and CPU power are clearly the primary constraints for what can be done on older players.

Does anyone know of an accurate list of Sonos hardware models with the few technical/hardware details we know of?

I suggest you read the oft-referenced article from Monty Montgomery of Xiph (the developers of, amongst many projects, FLAC).

Sonos have stated a couple of years ago that, whilst they can see a slight case for 24-bit (and the current hardware could apparently support it in theory), higher sampling rates make no sense.
https://www.whathifi.com/features/sonos-plans-brighter-and-brighter-wireless-music-future : "the math just isn’t there"

Sonos only says that because they don't want to spend the money to improve their products. Pretty obvious.

It isn't obvious to me. Hi Res is just audiophile Kool Aid. If the Connect needs a refresh it is for adding other features to it than the Hi Res red herring.

And since audiophiles buy purely based on brand labels and price points, an Audiophile Sonos Connect is a contradiction in terms.

Sonos only says that because they don't want to spend the money to improve their products. Pretty obvious.

Pretty obvious that audiophools aren’t terribly bright.

Sonos only says that because they don't want to spend the money to improve their products. Pretty obvious.
Pretty obvious common sense. Point me to a controlled double-blind test which consistently proves that hi-res is superior* and I'll rally to the cause.

(* Note: "superior" not "different". It's well established that ultrasonics can degrade the sound, due to distortion introduced by downstream equipment.)

Point me to a controlled double-blind test
Audiophiles trump this argument by questioning the validity of double blind testing in the audio domain. At that point there is nothing more to be said to bridge an unbridgeable chasm of belief systems.

Audiophiles trump this argument by questioning the validity of double blind testing in the audio domain.
In the absence of evidence the decision must be based on pure faith. Thankfully Sonos is still an engineering company.