The linked example using standard mic setups from that room and similar situations elsewhere sound good, so that's the obvious and simple path to take for this specific situation and a good starting point.
But furthering the more general discussion of far off-center location..
When I'm in similar off center positions taping off a stereo (or 2 stack array) I use A/B and choose between cards and hypers based on distance. With both mics mounted parallel on the bar, 6" or so apart, I aim them at the center of the stage. By doing this both mics will get the same basic signal from primarily the closest stack. But, because of the slight time delay between the mics, there is just enough of a difference between channels to give a limited illusion of stereo, not ideal but better than just pointing directly at one stack, IMO.
Are you keeping the bar itself parallel to the edge of the stage, or rotating the whole thing to aim at the center of the stage?
A parallel spaced microphone arrangement is a good approach, using time-of-arrival rather than signal level for stereo difference when the directional balance at the recording position is askew.
Combining a few of these ideas so they work together, here's an approach I'd like to hear-
>Use an AB spaced parallel microphone arrangement.
>Orient the AB array axis with respect to the apparent acoustic center (mic-bar perpendicular to a line pointing at that apparent acoustic center - the point it with closed eyes thing).
>Use very directional mics like hypers or figure-8's and
point them at the PA on the opposite side (cardioids and probably supercards will not have enough off-axis sensitivity difference across their forward hemisphere for the near-side/far-side balancing part of this to work).
You then have a sort of AB parallelogram microphone arrangement. The mics remain parallel to each other, but are no longer perpendicular to the mic bar. The idea is that we are no longer using signal level difference based on angle difference between mics for stereo image (AB), but are using the microphone's pattern sensitivity to reduce the level of pickup from the near-side PA relative to the on-stage sound and far-side PA. That can only work if the two mics are oriented parallel to each other and directional enough that the level differences
arriving from different angles across the stage are somewhat reduced by the near-side PA being off-axis to both mics and the far-side PA being on axis to both mics. It's a special PAS condition, pointing at the opposite side.
>We get stereo interest and sense of space from the AB spacing.
>The image remains centered because the energy in each channel is about the same (the pointing the whole array at the apparent acoustic center thing).
>We get a more balanced pickup of sources in the room by reducing pickup sensitivity for the close sources which are more off-axis, and increasing pickup sensitivity towards the far-side sources which are on-axis, based on the parallel arrangement of directional mics.
>We get some corrective image balance by effectively having sound from the far side arriving at the left mic prior to its arrival at the right mic, while the sound from the close-side arrives simultaneously at both mics. The left mic is slightly forward of the other mic when viewed from the far side. (This is a Michael Williams MultiMicrophoneArrayDesign MMAD principle)
Here's a plan-view sketch, looking down from above-