Hmmm: I would have said that you were right and that your taper friend was wrong. As you pointed out, the angular width of the sound sources is often relatively small in the type of recording that we're talking about. But narrowing the angle between microphones in a coincident or closely-spaced pair doesn't increase the proportion of direct sound in the recording by very much. To a much greater extent, rather, it makes the two channels of the recording much more similar to one another than they were before--which pushes the recording toward mono.
A more appropriate response, if it's possible, might be to put more distance between the microphones, to use microphones with greater directivity (if you're not using supercardioids already), and/or to use microphones with a moderate upper-midrange elevation.
Exactly these ingredients in combination..
Let me try and clarify why I came to understand how my taper friend was correct in a vitally important way, at least with respect to something more fundamental than what I was trying to explain to him at the time. Apologies in advance for what may be a long post. I acknowledge my susceptibility to identity politics / niche-seeking to the same degree as anyone else here at TS, although I try to I approach it as more of a high level gestalt engineering exercise applied to this unusual type of hobbyist music recording. A type of recording which has not been rigorously explored in the same way more typical professional methods of recording have. And because of that I especially value alternate points of view on any of this!
My current understanding is that both my friend and myself were correct, albeit about different things, and he was getting at something more fundamental with regards to recording a PA amplified band from an audience position than I. My realization that followed was about determining the appropriate hierarchy for prioritizing these things and addressing them in a logical order.
There is no substitute for an appropriate recording position, yet we can still seek to make the best of a compromised situation. The questions are: To what degree can we compensate at all? What is the most important thing to optimize? What can we actually do about it? Then, having identified and addressed the most important aspect, what's the next most important aspect and what can be done to improve that without messing up the more-essential thing?
A pair of microphones is only able to sample the sound field which exists at that location. As such, a recording made from an acoustically distant location cannot escape being dominated by reverberant far-field sound, regardless of the microphone technique used. Unless we can mix in a feed from the SBD, which is the addition of content from microphones placed close to the source(s), we are more or less stuck with the basic sound balance present at the distant recording position.
In such a situation it is advantageous to arrange things so as to get the sound of interest as clear and proximate as possible. It's difficult to enjoy a recording when one can't make out the lyrics or hear the musical phrases clearly. Those things are more important than good stereo aspects. A clear mono recording is more enjoyable than an otherwise unintelligible stereo recording*. Clarity and proximity are more fundamentally important than stereo aspects.
Let's start by reducing the problem to making a mono recording using a single microphone and recording channel. A microphone with a moderate upper-midrange elevation helps with clarity when recording at a distance, so let's start with that. Next what pickup pattern is appropriate? Say I choose a highly directional microphone like a quality supercardioid which, on average, favors sound arriving from the direction in which I point it over all other directions. In which direction should I point it? Will I hear a difference if the single directional microphone is pointed away from the stage and towards the back wall, towards a side wall, or toward the stage and PA sources? Its pretty self-evident that I want to point the microphone in the direction of the sound source regardless of my distance from it. That helps because even though the
level of sound arriving from any direction is essentially the same as any other due to the distant recording position, the
quality of sound arriving from the direction of the source will be somewhat more clear than that arriving from other directions. Not necessarily by very much (a consequence of being far away), yet enough to make pointing the microphone toward the source a meaningful improvement. If anyone really does doubts this, I can provide samples of identical supercardioid microphones pointed directly toward the stage, directly away from the stage, and directly toward the sidewalls, all made simultaneously from a distant recording location. The differences in most cases are easily audible, although sometimes not as significant as many may expect. Otherwise I may as well use a less directional microphone that is likely to have smoother off-axis response, a more extended low frequency response, and is less susceptible to handling/wind noise like say an omni. Which I would if I could only place the microphone close enough.
For the same reason, It's also important that the directional microphone I use have a well-behaved off-axis response that differs from it's on-axis response ideally only in level, but not significantly in frequency response. Because again, the soundfield energy at that distant location is arriving with pretty much equal level from all directions, so the microphone really needs to sound good across all angles of arrival.
Okay, I've now addressed the most significant issue as best I can, by pointing a well behaved directional microphone at the source. Next, how might I introduce a second microphone channel to make this into a suitable stereo configuration without overly compromising the more important thing I've just achieved?
I can add a second supercardioid facing directly toward the stage, AB spaced from the first like I would arrange a pair omnis. This produces time of arrival stereo like a pair of omnis without compromise to the first aspect. But because ensemble on stage fills some small angle as viewed from my distant recording location, and because the PA speakers are a major contributor of the sound and are typically placed on either side of the stage, I choose to point the Left supercardioid directly at left PA speaker and the Right supercard at the right PA speaker. Now both microphones remain on axis with the primary sources, but now have some angle between them, as determined by the width between PA speakers and my distance from them.
Once I know that angle, I can determine the optimal spacing between the two microphones based on the collective work of Williams, Sengpiel, Wittek, etc. which has been made easier by referring to the Improved PAS table. This is also good because any angle between microphones will reduce the need for as much spacing between them, making the setup more practical while at the same time introducing some welcome degree of of level-difference stereo in addition to time-of-arrival stereo.