There is a german psychologists study of 3 audio systems blind tested to more than 50 members of the public. Very interesting, the trouble is I can't remember which web site I read it on. Might be linked through audio asylum, or audionove.
Blind tests are complex, not just because some of the differences are percieved over time, and in ways which are difficult if not impossible to articulate.
As far as your scepticism of some refinements, I was too, but you can hear differences clearly if you exagerate. If you have biwireable speakers, wire the bass with ordinary shopping mall quality wire, and the highs with a decent audiograde. Then reverse the two. The difference should be easily heard.