I am of the opinion that the only wrong way to live one's life is to insist that others are wrong, especially if I insist I'm right. I tend to take everything with a grain of salt until I've had a chance to knock on theory AND practice quite a bit myself, and understand the benefits or shortcomings of whatever I am actively trying to do. Noah, Gutbucket, and DSatz have all given me more endless practical knowledge that I know what to do with, and I've been spending at least the past 5+ years of my life collectively unpacking their wisdom and trying it for myself.
I tend to run 3 pairs these days because I can. When it's inconvenient for me to run that third pair OR if I'm in a situation where I don't think the third pair will give me any great benefit, I only run two pairs. I never run one pair on its own, it's always "easy enough" for me to run the second pair and get a different recording to experiment with and turn taping into a science laboratory.
There are two extremes of stereo microphone techniques: coincident microphones (XY, Mid-Side, ambisonics) and spaced AB pairs. These two extremes correlate with the two most well known (but by no means complete, exhaustive, or universal) binaural cues that we use for localization and perception of sound in space: the difference in level cue, and the difference in time cue.
The idea behind the "classic" stereo near-coincident techniques such as ORTF, DIN, NOS, etc is to get a blend of these two cues in a manner that the engineer designing them finds both pleasant to listen to on playback, and accurate for depicting the acoustic field that the microphones captured. We use them as a template, and I now notice a larger number of tapers going "off script" when we find techniques we enjoy better - I personally prefer the bass response and increased time cue of NOS's wider spacing, but narrow the mic angle to get more of the image "in front" of my recording and to get closer to on-axis with the stacks when taping.
I echo Gutbucket's sentiments that many tapers likely shy away from multi mic techniques because it's possible to get into trouble pretty quickly - not just from combs, but from phase misalignment. Trying to mix two near-coincident pairs is difficult to get right in my experience, but that's not to say one can't get reliably pleasant results with it - just that it's a problem I haven't figured out and haven't bothered trying to solve.
For multi mic pairs, I have had the most success running AB Omnis or sub cards, and mid-side or XY pairs between them, then mixing down in post. You can stream the results of this method using 60cm AB mk21s and a mk4/mk8 mid-side pair here:
http://phish.in/2019-08-30 if I use my Omnis today for the AB pair and add a third pair of hypers, cardioids, or subcards in a classic near-coincident technique for consistency and because I like the results, I'm at 3 pairs right there. Given I own 4 pairs of microphones, I've been thinking long-term about switching to a deck that can run all 4 pairs when I want to... but time and my finances will dictate if I ever get around to that.
While the basic concept of comb filtering is straightforward enough, we are still receiving
many, many, many combs from a variety of sources. As the pinna (outer earlobe) and concha (inner bowl surrounding your ear canal) serve to funnel sound, they also serve to provide a quick reflection off a certain point of your ear, and this contributes strongly to your sense of elevation or height of an audio source. Likewise, there is a comb between your left and right ear in the free field, not to mention all of the combs that are a result of arriving specular reflections in any acoustic space with a boundary capable of reflecting sound. David Griesinger is a close friend, and former mentor and boss of mine - he's a wonderful and smart individual. I've discussed live music taping with him a number of times, he thinks we are simultaneously noble and crazy for attempting to record amplified pop music in a noisy crowd. He himself is an extremely accomplished classical recording engineer, and his go-tos are binaural microphones for recordings he makes in Boston Symphony Hall; when he hangs main pairs, he's a big fan of hypercardioid SDCs in ORTF-ish, though he doesn't do that any more the last I spoke with him.