Here's a link to the explanation about the tests performed by Meyer and Moran that were submitted in the recent AES paper, "Audibility of a CD-Standard A/D/A Loop Inserted into High-Resolution Audio Playback."
It seems like a fair test, even though the original home audio system was by no means up to the task. They eventually went to a few mixing and mastering facilities and tested some young ears. Their findings indicate that there were no significant differences between the SACD layer and the downsampled conversion except at levels higher than "normal." Normal was considered to be 85 dB at the listening chair, calibrated using a pink noise sample. Theoretically, some quieter jazz and classical music would require the 10 dB boost they used to hear the details, so it's halfway feasible that a normal user might hear differences in low level accuracy.
Is it worth paying more for a slightly higher fidelity recording with copy protection? Maybe not. People like Head-Fier markl have noticed significant quality differences between early CDs (esp. from places like West Germany) and current ones, so perhaps it would be a more cost-effective idea to focus high-res efforts on a better CD mastering process, like those employed during the manufacture of HDCD, XRCD, XRCD2, K2, and K2HD.
True, the new cd's are all hot mastered, so they blow your head off.
I also noticed huge differences in quality of cd's. Almost all new popular cd's are way to hard mastered, they blow your ears off and sound distorted.
Good old recordings however sound wonderfull.
Between the cd and xrcd of money for nothing from the dire straits is a significant difference beneficial to xrcd.
Also people have found differences between sacd recordings; some are as bad or good as cd. Some are better then cd but that's all in the mastering, not the format, wich has the 6 bit highs limitation!
It's all in the mastering of the recording.