...but seriously...
The problem with these questions and these analyses is that we're not answering the question that is being asked, 'what sounds better?"
Having worked with engineers for all of my professional career, representing the exponent of
humanity (my customer) in a lot of conversations like this, it has been my experience that engineers will often answer the question they think you should have asked, rather than the question that was actually asked. They do this because, subconsciously, they know that their fancy maths cannot answer your "irrational" question, and so, try to distill (digitize?) your question into something to which they can apply their formulas and knowledge. Or, quite possibly, it is because they are devoid of all humanity in the first place, and a question like the above "DOES NOT COMPUTE."
As a result, engineers hear the question "WHAT SOUNDS BETTER" and, after fancy maths and slideshows, video tutorials and experiments, come up with the answer to the question, "WHICH SOUND IS BETTER." Thanks for all of the work, but that's not the question we asked. Because we're dealing with a human emotion here, and have no variable for this human factor included, the answer works for some, but leaves the vast majority behind. That human factor is so complex, maybe even absurd, certainly unknowable, but to ignore it is akin to applying Newtonian physics to the quantum interactions of the wider Universe. T'ain't gonna work.
Now I'm certainly biased -since we live in an ANALOG world, I honestly believe that an ANALOG representation of that world, no matter how flawed or restricted it may be, is "better" than a DIGITAL sample, no matter how close to "perfect" it may be. But that's not based on math (or even "facts") so really only applies to my own personal universe, and valid only within the crusty and polluted confines of my own head.
So what. Big deal.