In a recent issue of The Absolute Sound magazine Alan Taffel compared SACDs with their equivalent DSD downloads. In the case of Rickie Lee Jones' Traffic From Paradise Taffel preferred the SACD, and concluded that the SACD was "better sounding." While he was right about hearing differences, he was less right in attributing the fault to the DSD file. In point of fact the two format's files were bit-perfect identical according to Gus Skinnas, who was directly involved in the mastering process.
In a subsequent follow-up post Taffel explained that after talking with Skinnas he realized that the differences he heard were being caused by the differences in the signal path between his SACD player and his computer audio system. According to Taffel, "In the case of the hi-res file, the DSD bits flowed over USB. However, when playing the SACD those same bits traversed CH Precision's proprietary CH-Link interface."
And what can we take away from this? First, the obvious stuff - everything in the signal chain contributes to sound quality. Perhaps a different cable or USB to S/PDIF convertor and the DSD Download file might have sounded better than the SACD. If I were in Taffel's position I'd try swapping some components in the signal chain for others just to hear what happens.
Computer audio is a strange and wonderful world where seemingly inconsequential details such as cable routing and grounding schemes can have major effects on overall sound quality, which makes it all the more difficult to generalize and theorize about performance limiting causes and solutions.
Ideally, if the SACD player and the computer audio front end were both performing optimally and feeding the same DAC, there should be no differences in sound quality between the two sources while playing back the same digital file. But obviously, we aren't there yet, are we?