Is that streaming from a Plex server? If so, is anything happening there? 

I suspect the Plexamp software is playing a role in the issue mentioned here, as it’s description on the iOS App Store says it has in-built loudness levelling, a configurable pre-amp and parametric EQ. 
 

I don’t have a plex server setup here to go onto try out the software, but I suspect that is perhaps what might be at the root of the difference being heard in the audio here.
 

If the plexamp software suits, then the answer has to be to stick with what you prefer.

Looked at the settings: 

loudness levelling on.

EQ switched off.

Preamp at +4db

Server is on a 10th Core i7 Intel NUC

I’m definitely sticking with this configuration, especially as the App sticks to the Amp like glue; no annoying casting disconnects, restarts etc.

It does make me wonder if all streamers are not the same though.

… until your phone gets distracted, or removed. Looping a stream out to a mobile and back again always carries risks. 

What’s that in detail? Is it per-track gain adjustment based on data in the tags, or is it on-the-fly dynamic range compression? The latter could have a marked effect on the sound. 

Evidently it carries out an analysis and marks each item with the dB gain in the Plex library, and then applies a volume adjustment when you stream.

So it just adjusts the volume. It’s on, and it surely won’t make it sound better. It certainly doesn’t sound louder when I do an AB, volume appears to be very similar with each source. Louder is often construed as better so it can’t be that.

On the streaming issue via the iPhone, it is excellent. I think Plex gave it some attention recently. It doesn’t get distracted. If you close Plexamp, then reopen later it is connected to the Sonos Amp as before. 

I switched from a connect to a port. I’ve got it feeding a rogue sphinx integrated amp and floor standing speakers. I thought about other alternatives (blue node), but I have a lot of sonos stuff in the house and did not want to deal with screams from the family. 

My perception is that the port was an upgrade from the connect on its own, sq wise. I’ve also experimented a bit and found that:

1. Having it feed a good external DAC, does make a difference. 
2. Swapping out the power supply for an ifi power supply that reduces the noise of the switching power supply also made a difference, and is a cheap upgrade for about $50. More dynamic range. Better base. 
3. If you are feeding an external dac (I’m using an Ares II) using an iFi Spdif purifier 2 made a noticeable difference. I was surprised. I think, however that  will be highly dependent on the DAC, and how well they filter and reclock the signal they get from the port.