This is the case more often than you would think. I understand your hesitation to believe this coming from your background, but there’s a few extra variables in this world (as I’m sure there is in enterprise) that add additional confounds. The endpoint audio devices often feature atrocious WiFi hardware and software stacks. Various combinations of low gain antennas, poor RF engineering, subpar drivers, and idiosyncratic buffering techniques plague the industry. Things are slowly getting better, but existing devices aren’t going anywhere for the time being.
To back up my claims about urban environments, I’ve attached a speedtest I ran on my Amazon Fire Tablet using the 2.4GHz network. I’m about ~30ft from the router and there’s two walls blocking LOS but both doors are open:
Fast enough to stream high-res FLAC? Yes. But what happens when someone else in the household decides to fire up a 1080P Netflix or YouTube stream @ 6 mbps? Or what happens when you use Roon, which decodes all audio to raw PCM before transmitting over the network? That’s 6 mbps for a 24/96 stream, this connection would choke for 24/192 RAAT.
What if you want to do multizone streaming? Now you need to double the bandwidth requirements for each device.
The next question may be why don’t you just stream the CD quality audio or take in the hi-res and downsample?
Hosting multiple different formats of the same content is confusing, leads to extra complexity in the UI, and increases storage requirements.
Downsampling on device takes extra CPU horsepower and can’t be relied upon in a multi-vendor situation. It adds latency, is frowned upon by people who care about sound quality, and adds an extra layer of complexity.
Unfortunately, none of this is as easy as it seems.
For example they could troubleshoot and upgrade, or they could be satisfied with 320Mp3/AAC.
You’re right, they could. MQA is betting that there’s enough people out there who would rather not do either of those things that it will help drive adoption for them. Time will tell.