The problem with those listening tests is the playback medium is not controlled: ie, what were they using (what headphones, what headphone amp, what DAC, etc; or what speakers, what speaker amp, what DAC, was the room properly treated, etc); and what is their level of hearing damage (ie, their age)?

If everyone is taking their test on their phone with low grade earbuds plugged in, or exceptionally bad car audio systems, everything sounds bad, ergo, the same.

I own a pair of Stax, among other higher end headphones, with the proper gear to drive them. I can tell the difference between modern-LAME-encoded 320kbps MP3 and lossless, albeit just barely, in an ABX test. Everything 64kbps sounds bad unless it's something like Opus doing conventional speech.

I know I'm setting the bar high, but I don't think I even have "golden ears", I think anyone can reasonably hear the difference, they just need to quit bothering with low end gear.

I mean, people just on HN buy Tesla Model Ses for $100k+ without even thinking twice, I find it hard to believe they won't also drop $1k on headphones and headphone gear at the very minimum if they care about audio.