If you ever have the opportunity, the result from such a blinded test would be fascinating. Of course, it would also be interesting to compare the actual sound waves from each source, to see if there was a physical difference. If they're identical, and yet your blinded test still shows that you can differentiate... well, presumably there's some aspect to the sound that isn't properly being measured.
And of course the opposite is even more telling, if the soundwaves are measurably different, but the blinded test reveals that difference cannot be detected by humans.
There is obviously a point of resolution at which a digital sine wave is indistinguishable from its analogue, by human ears, and that will come long before it cannot be measured by machine.
I find it a plausible possibility that some poople might be able to tell the difference, since the general market is only driven to provide sufficient quality for most people. But still, considering you just add one bit and you double the quality, it seems increasingly unlikely that this kind of thing is a problem. At any rate a double-blind test is the only real way to sort out the men from the boys.
A good test might be to record a vinyl record onto CD and play them back blind.