I have the Wilson Audio LP Don and Peter referred to. Its purpose, as stated on the jacket notes, was:
... to produce a test recording of musical material to specifically demonstrate one test variable ... one generation of digital processing by a professional grade digital recorder. Any difference in sound between side one ("Original") and side two ("Digital") can only be caused by distortions introduced by the digital processing itself. All other excuses used by digital apologists to explain less than perfect sound from digital ... all have been cancelled out."
The LP was produced using very high quality material from two Wilson Audio recordings, one of ragtime piano and one of a jazz trio. They had been recorded and were played back during the mastering using Dave Wilson's custom 30 ips 1/2 inch analog recorder. Both masters were cut within one hour of each other using the same equipment, with levels carefully matched. The only difference between the two sides is that on the "digital" side the Soundstream 50 kHz 16 bit digital recorder, which as I recall was state of the art in its day, was inserted into the signal path between the analog tape machine and the mastering equipment.
That methodology made the question of which side was "better" irrelevant. The only question was the existence or non-existence of differences between the two sides. And exist they did, the differences being clearly audible, as I recall, even on the fairly modest system (by high end standards) that I was using in the mid-1980s.
Thirty or so years later it seems like a re-do of that kind of comparison would very much be of interest, even though I whole-heartedly agree with Mapman that "either technology might win anytime in any particular case." I will be signing up and contributing my $10.
Thanks, Peter, for calling this to our attention.
Regards,
-- Al