I do respect what you're saying--as someone said, if something sounds good, it's a good recording (implication: regardless of how it was made). Some very mysteriously spooky wonderful old mono classical recordings have been made with microphones that had a fairly high amount of distortion. When you work with a certain set of gear long enough and really listen to what you're doing, you learn how to bring out its potential even though more accurate equipment has greater potential in more situations--but maybe for that one recording, everything (including objective shortcomings of method and/or equipment) "fell into place". Some amounts and kinds of distortion under the right circumstances can sound awfully good.
Hell, the entire LP recording and manufacturing process is so full of distortion-causing steps that have no counterpart in the CD mastering process, CDs made from LP master tapes definitely sound different from the corresponding LPs as people quickly discovered, and having heard some of those LP master tapes directly when I worked at RCA, I know that at least some of them actually sounded quite crappy. But over the decades, the people working at all the different levels of production had learned how to make fantastic-sounding LPs, even if that required making master tapes that sounded distinctly bad. Certain things along the way compensated for certain other things if you knew what you were doing. There's a persistent story that someone accidentally dropped a cheese sandwich into one batch of molten vinyl at a pressing plant and the resulting vinyl was quieter than usual, so cheese was added from then on--I doubt that it's true, but it COULD be true--things do get discovered that way sometimes, and that's the point, I think. If cheese makes vinyl quieter, by all means add cheese. But then maybe see if you can figure out why that works, because even further improvement may be possible. Plus, you might find out that it was actually the mustard that did it (in the parlor, with a candlestick).
Schoeps used to make a kind of miniature stereo mike called the CMXY 4 V, with a pair of CCM 4 V microphones in modified housings side by side with only a very narrow gap between them. Each miniature microphone was mounted on a toothed gear, engaged so that the two of them always rotated in contrary motion. You could set any angle you wanted between them, vary that angle by rotating one mike, and the other one would rotate the opposite way so that the center line between them would stay put. They were of course measured that way and the interference between the two adjacent capsules could easily be seen--but it sounded pretty good nonetheless, in most people's opinion.
I saw this arrangement and at the time, was neither a fan of the MK 4 V nor of coincident stereo recording with cardioids. I asked them to make up a similar microphone with supercardioids instead, although I had never used the MK 41 V before. They kindly made one and lent it to me for testing. I recorded a concert (chamber music at the Isabella Stewart Gardner Museum in Boston as I recall), and was very disappointed in the sound. It was less clear than I had expected, and generally unattractive sounding. Schoeps (again kindly) took the microphone back--they were interested, but somehow not too surprised by my reaction. This experience soured me on the MK 41 V for some years--until I bought another pair and placed them more than a capsule's width apart and facing outward for stereo. Only then did I realize that I actually preferred their sound over the forward-facing MK 41 which I had used probably 1,000 times. Also, a contact at Schoeps revealed to me that another customer before me had requested the same kind of stereo mike with supercardioids, had a similar reaction to its sound quality, and returned it. The mike that I tried may even have been the same one, I don't know.
I came away with the guess that the closer a transducer is to having a pure pressure-gradient response, the more vulnerable it may be to this type of interference. That makes sense to me according to my notions of how things work; I haven't tried to prove or disprove it. If it's right, though, figure-8 microphones would be the most vulnerable, and that's how come this question interested me so much. I do have some direct experience with one of the issues that I'm raising, in other words; my questions aren't based on theory alone, although the theory is quite clear. The interference and lack of coincidence may well matter less than other things such as very astute placement of good microphones in good-sounding halls with good musicians playing in them--but that doesn't mean that they don't matter at all; people would have to experiment and listen to find that out specifically.
The thing is, maybe everyone would prefer the version that goes against the theory. That has to be allowed for, although then I would want to dig deeper. It may be, as with the good-sounding LPs made from crappy-sounding master tapes, that a defect in recording technique is compensating for some equal and opposite defect in the equipment or the acoustical situation. Just because adding cheese to vinyl made the vinyl quieter (if the story is true) doesn't mean that cheese should be added to everything.