Like most discussions of ABX testing, particularly the ones where it it is reified and turned in to the holy grail of science, a very important point about using science to make statements about the nature of reality is missed. This is the tension between reaching conclusions through naturalistic observation vs. laboratory experimentation. The natural observer is able to view the phenomenon in relatively unadulterated form, but has no control over the circumstances involved. This leads to conclusions that pertain directly to a real-life phenomenon, but are hard to generalize beyond that one observation as there is no control over the variables being assessed, and potential confounds abound.
So the solution is to take it to the lab, right? Where the situation can be completely controlled?
Not so fast. This brings in a whole new set of pitfalls. It is an entirely synthetic setting. By eliminating potential confounds, it also eliminates a whole host of other factors that are always present when the thing under study takes place in the wild. It is almost impossible to measure something without changing it, let alone manipulate it in an experimental design. So in this case, the threat to generalization beyond the lab, and hence a fuller understanding of what you are studying, is the lack of similarity to real life.
So what does this navel-gazing about the scientific method have to do with ABX testing? Simple. It simply has no relevance to how people actually listen to music. At least I hope it doesn't. Therefore its results can provide information, but are hardly the last word on how a component presents itself for its intended purpose, which is not being a part of a lab experiment.
So I don't believe ABX or any other type of lab experimentation should be abandoned, but it has limited applicability to truly understanding the complex psychoacoustic experience of enjoying music. Both objective and subjective perceptions have their place. But there appears to be a difference in how they are discussed. Subjectivists almost always give some version of YMMV. ABXers say that if they can't hear it, neither can you.
One reason it would have been nice to see some discussion in this review about how well the Xonar bitstreams the hi-res codecs is that it is currently the only PC component that can do so. Decoding them is quite possibly tantamount to unpacking a RAR file, but there is not yet a consensus on this.
Finally, Devonvar, I can't let your conclusions about vinyl pass without comment. To wit
In fact, I would argue that one of the reasons vinyl remains so popular is because it blurs the audio in a way that hides flaws in the mix and mastering process.
This has simply not been my experience with any turntable Rega-entry-level or better. Aside from surface noise, the main difference I have observed between good vinyl and good digital is that vinyl conveys the timbres and textures of instruments more realistically. It doesn't blur anything, which can be unfortunate in the case of bad recordings.
MikeC, IIRC you have some experience with high-end turntables. Care to weigh in?