Secondly: lossy, audio data compression is not the same as audio dynamic range compression; they are only related inasmuch as audio with less dynamic range can be compressed into smaller files. You could have 24-bit FLAC files that would still sound shitty if they had been mastered with too much dynamic range compression.
Thirdly, and most importantly: I think there is a difference between music as art, which has a relatively small audience, and music as a consumable item, which has a big audience. The artists interviewed could maybe insist on only the best quality versions of their music being available. They could say "sayonara" to Spotify, "arrivederci" to Apple and "piss off" to Pandora. In doing so they would sacrifice a large part of their audience and their business. Would they do that?
There's no question that the problems highlighted are eminently solveable, now or soon. It's also clear that they won't be until there is a demand. The film essentially claims that most people don't know to ask. I think the better question is whether they would care if they did.
[Asides: David Hepworth makes similar points better than I can. And Greg Milner's book "Perfecting Sound Forever" is a fascinating insight into the history, techniques, technicalities and politics of recorded sound.]