We all know recording with 24 bits has tangible benefts. This test was about delivery; SACD could not be distinguished form 16/44. 16/44 has all the resolution we need for playback. Using 24/96 or whatever is perfectly legal and usefull (at least for recording), but claiming a big, audible difference in playback quality is simply not true.
Let's not mix up two different things, recording and delivery.
-----------------
Funny note: on a hifi bulletin board there has been two long winding threads about audible audio quality: one about different MP3 versions, one about this same CD versus ri-res formats. The funny part is that MP3 discussion is quite factual, people calmly admit they can not tell the least compressed MP3s apart, or distinguish them form CD quality. The CD/hi-res thread is much more emotional, there the whole question of even testing such a self evident fact [that hi-res is superior] is questioned, the validity of making test files is questioned and people refuse to take part in the test (or at least they decline to make any comments about their personal results).
I think this only can be caused by the monetary sacrafices these people have made. You can test different MP3 compressions without investing any money; any digital player can play anything MP3 and CD-WAV, so you are not going to loose face by admitting the best compressions sound the same. Not so with SACD/hi-res. Some people have invested thousands to play these formats, they will not admit they can not hear any improvement, they even refuse to listen to a simple test file to see it the difference can be heard.