One of the problems I have with A/B and ABX tests (I've been in a boatload when I worked in the industry, but never for headphones) is that they don't allow me to get familiarized enough with the sound signatures on high quality gear to actually figure out what I preferred over the other. Differences could certainly be heard fairly quickly, but for equipment at similar levels of performance but with different sound signatures - it hasn't been until I'd heard them extensively using all kinds of different music during different times of the day, over different days, etc, that I can truly decide which I prefer. This implies, at least in my case, that the reference frame for 'maximum enjoyment value' is dependent on many factors. Some of these, like soundstage and positioning are not measurable, and the effect that these have on the perceived level of detailing is subsequently lost to us still. Others, such as the impact of activities earlier in the day vs impressions of different sound signatures, or even the order in which the sound signatures are presented - are completely lost on us still.
I'm all for quantitative measurements, but they simply do not capture everything that our brains register. You don't have to go far into behavioral and neural psychology before you realize how little we can control the subjective experiences we have (and appreciating musical reproduction is inherently subjective). I definitely haven't seen any neural psychology results mapped to the people doing A/B or ABX tests, and before we can connect many more scientific disciplines - we're still way too crude to call A/B or ABX fully conclusive. Heck - we need to wait until neural psychology has a better grasp of how different individuals brains work when reaching the same conclusion before we can go into why this may be so, and if behavioral psychology factors are impacting this, and only then... do we get electronics in to try and come to any definitive yes or no answer to this question.
I won't even get started on the notion of 'is there a universal truth', as the only way to answer that is to get into the insanely deep philosophical debate that so many have spent their entire lives at trying to nail down. In fact, the only way to answer this fully assumes that everyone can accept that there is one fully true answer, which obviously is an inherent problem with the question itself. Crudely, this implies that nobody is in a position to argue that Z is conclusively true because X and Y are known. All we are doing is inferring that this is likely to be true. In other words, we might just not know that pigs really do fly, but that none of us have been watching when they did. Vice versa, it may very well be that Z is conclusively true because of X and Y as the reason we haven't seen pigs fly is that there really isn't any additional magic aspect to this.
With this post, I hope all believers in a specific test as conclusive start accepting that the tests themselves are designed based on a very limited understanding of the human perception and what affects it. What goes into defining a test like this is really best guesses on a limited number of variables, while all other variables that couldn't be controlled are simply ignored. I also hope that all who believe there is much more to cables than we already know, that my argument for additional variables is ONLY an argument for the testing procedures as inherently flawed. They are still the best repeatable test cases for a large number of people to draw generalized conclusions. They may differ significantly from how I come to terms with what I prefer, but the thread wasn't about how I do things so I leave that unsaid.
Oh, and I hope some of you feel that I must be a very confused person. That part, I can confirm.