fr0g said:
They fail because there is no audible difference. Unless you happen to have the hearing of a mosquito. 44 KHz will give you a perfect analogue reconstruction up to 22 KHz. No human can hear that high frequency. So, it explains nothing of the sort.
Actually you're clearly wrong without a doubt. You cannot say logically that you know beyond any exclusion or exception that "there is no audible difference". You can only say that you haven't heard it, or what tests you've read about don't report it. Too many people make absolutist statements that fail on their face, and you're doing it here.
Edit: I should add that's there's a world beyond frequency - phase and impulse, etc. The 44 khz of the CD/WAV track isn't the same as 44 khz in analog. The WAV track is a sampling in bits (1408 kilobits/second), and the bits can't account for everything on the extremely complex analog sound waves, since those waves aren't single tones, and they're often of a complexity such that only highly trained people can hear certain things in those sounds in the analog domain, before it even gets to digital.
So for people who can't hear differences that are easily measured - great!! Save your money and avoid building a large pricy collection of 96k or 192k or (gasp) 384k tracks. But when you build an expensive collection of 44 khz tracks and then discover later that you can hear the difference after all, it could be a serious and expensive disappointment.