A core issue in the design of audio equipment is, "what really matters?"
Not everything that can readily be measured matters. For instance, total harmonic distortion measurements have essentially no correlation with perceived sound quality, but certain mathematical weightings of THD measurements do.
Years ago as an amateur speaker designer I decided to build a speaker that attained the Holy Grail - namely, time and phase coherence (true first order crossover with drivers aligned on a sloped baffle) along with flat frequency response (accomplished over several days of crossover refinement). As I got closer and closer to "flat" response, the speakers sounded worse and worse. But I pressed on, unquenchable in my faith that when I got to the Promised Land, the Angels would Sing. Finally, I achieved the impossible: A two-way loudspeaker that measured plus or minus .75 dB from 45 Hz to 10 kHz, 1/3 octave pink noise (it went higher, but I didn't trust my measurements up there, and in retrospect probably shouldn't have trusted them under 200 Hz). How did it sound? Terrible! Possibly the worst sounding speaker I had ever made. The imaging was holographic, though.
Now, since then I've heard loudspeakers that measure very close to flat that sound good, and others that sound awful. So I can't reliably say that flat response sounds good or bad. And, which "flat" measurement are we talking about? On-axis, 30 degree listening window, front hemispherical, omnidirectional (power response), in-room or anechoic, and at what distance?
Time and phase response is an issue I read about quite a bit in the Journal of the Audio Engineering Society years ago. My recollection is that, at that time, the research did not support the thesis that time and phase coherence was audibly significant on music program material. I don't think it's a bad thing, and I do think it helps imaging and probably dynamics, but a lot of experienced audio engineers don't think it's worth the compromises imposed.
I have my pet theories as to what characteristics and measurements correlate well with perceived sound quality (and my tomorrow's theory may be different from my today's). David Wilson has his, and Richard Hardesty has his. As an aside, who's to say that the exquisite craftsmanship of the Wilsons doesn't contribute to their enjoyment? In the restaurant industry, "presentation" (how the food looks) is an important issue. If it tastes like steak but looks like a dog t#rd, some people are going to be put off by it. (If the price is right, I might not care what it looks like; I've been known to have a hard time enjoying a first-class steak thinking of how much it costs - why, you could buy a set of tip-toes for that!!).
But back to "what really matters". I'm not convinced that David Wilson is clueless on that front, nor am I convinced that Richard Hardesty is. I think that in many cases advances in the area of understanding "what really matters" have not kept pace with advances in technology, so that in and of itself advanced technology doesn't promise any better results. My recent association with an industry professional who is doing core research into "what really matters" has convinced me that there is a great deal yet to be learned in this area, for it turns out that what the human hearing mechanism has a high tolerance for and what it has a low tolerance for does not neatly coincide with what is easily and commonly measured and/or calculated.
Duke