I'd also like to say that doing a blind test on music you've never heard is near useless. Just because one "sounds better" than the other doesn't mean it's better. The point of being an audiophile is to try and reproduce music (and other audio) exactly as it was recorded. Compression is known to "color" the sound.
For instance if the MP3 algorithm cut out some midbass making the lows and highs seem lower and higher, people might say it "sounds better". And you know what? It might! But that is NOT what was originally recorded and to me that is not being true to the artist(s) who made and mixed the program material. If artist "X" wanted it to sound a certain way, that's the way I want to hear it.
So to do a true test, use a CD that you know well. Make sure you've heard it on a decent system. Then make a MP3 from it and perform the blind test (also on a decent system). Personally I think a lot more people than you would expect will know the difference.
I was also wondering, why did they ever come out with SACD if no one could tell the difference? Why is DVD capable of better sound than CD if there is no difference?