Blind testing is good for the first two questions, but it's next to impossible to blind test headphones.

The question of preference, though, is what matters, and that's where I am now firmly convinced lossy compression is fine for me: the sort of differences that I could hear made zero difference to my enjoyment of actual music.