This isn't a comparison test. There's only one version uploaded. With only one version, it's hard to tell if many flaws you hear were introduced by sloppy uploading or if they were present in the master.
Yes! Which is why you won't hear me waste a whole lot of breath yammering about lossy compression.
I'd sure pay Spotify extra for lossless, because I'm weird in ways I'll reveal below, but I agree that people should give lossy compression a break. Well-encoded AAC and Vorbis averaging over 256 kbps are very transparent-sounding, in ways that never was possible with mp3. If I put in time, I get 5/6 right in this famous test from NPR's website[1], just because mp3 is awful and ancient.
But I double dare anyone to blind test Opus as low as ~128 and ~160 kbps and working upwards, with decent gear. Having grown up with shitty mp3s, it feels like magically good. And it's free software.
Even lossy-lossy transcoded AAC and Opus, which Youtube uses for a lot of stuff, sounds shockingly fine, most of the time, on most equipment, if the original copy was ok to begin with. All this is mostly passable, especially as background music.
That is, until you run into special circumstances, like listening attentively with halfway-decent equipment. Spotify's default normalization mode sometimes can adds dynamic compression, which sounds bad in itself. But this can make artifacts stand out in ways shouldn't (thank god for the 'quiet' setting, added sometime in 2018). This is especially true with poor source material, for example the stupidly bright 90s Led Zeppelin remasters, which still float around on tons of curated Spotify playlists, despite being superseded by really good releases.
So what I'm trying to say is that I want to maintain a music library I can pull up on any device and expect consistently good quality during playback. Take my little hobby I discussed here as an example (that is, me and my friends independently inventing the Japanese audiophile parlor/café concept) https://news.ycombinator.com/item?id=20583900
Just because I can't hear a guitar riff getting slightly distorted or hi-hats smeared when I listen at work, doesn't mean I won't hear it with in an acoustically outstanding room with 5k worth of audio gear. This problem is very pronounced with Youtube material when there's a poor supply chain, so I won't add a bunch of random garbage from Youtube in a playlist on the tram and expect to actually enjoy it later as I'm leaning back in a proper listening room.
Spotify, on the other hand is relatively close to providing a universally sane way to access music on any device.
Isn't that kind of the point? If you can't tell which is which without a visual cue (aka bias-generator) then they sound the same.