While Rock was posting I was editing my post above to split off the portion below as separate post, as it concerns the A-B hanging omnis approach. The intent is to clarify, make the more specifically relevant post above shorter, and talk about the details of an approach you may not care to pursue at this venue separately in this post. Feel free to ignore everything below if going the other route. I'm including this post for others who may be searching for info on installed recording setups since the wide omni A-B PA pseudo stack-tape approach works really well in some places.
In the photo you posted the PA split does look quite wide, making the hanging omni approach more challenging than in a smaller club. If going that route I'd want them closer than mirror ball distance to the stage, and not as wide spaced as the PA. More like halfway between stage and mirror ball, with each mic positioned at about the midway point of an imaginary line running between the mirror ball and the PA speaker. Maybe somewhat less spread than that, but want to make certain each microphone remains inside the coverage angle of the PA on it's side (Left mic inside the coverage pattern of the Left PA). Its okay to have each mic outside the coverage area of the PA speaker on the opposing side (positioned to the inside of the effective coverage cutoff angle of the PA on the opposite side), as that's essentially what happens in smaller clubs with omnis hanging in relative close proximity in front of either PA speaker anyway, which is known to work well from experience.
[full on speculation past this point, feel free to ignore] ..and such an arrangement may conceivably help with binaural-stereo translation of the direct-arriving PA sound by producing a stereo frequency response similar to that of a Jecklin disk. Geometry-wise it's totally different of course due to the wide vs narrow spacing. But a big part of the baffled stereo "thing" is the frequency dependent response the baffle imparts to sound arriving from the opposite side, which is emulated somewhat by being just outside of the effective coverage angle of the PA on the opposite side. Only in this case with some stereo hass delay to it in addition to the frequency shading. I think of 3 general scenarios with A-B omnis: 1) Typical A-B from the audience. Both microphones pickup both PA speakers, both direct-signal and direct-cross-signal has similar response, similar level, and a arrival time differences of no more than a few milliseconds. Diffuse reverberant pickup with similar arrival time differences; 2) Wide stack tape A-B with mics close in and directly in front of each widely separated PA speaker - each microphone dedicated to its own PA speaker. Direct signal has similar response, level and arrival time. Direct-cross-signal is negligible, no audibly significant direct sound cross signal between channels, any cross stereo delay is only from reverberant pickup, with maximally different arrival time difference; 3) Somewhere in between the previous two, as would be the case when setup as described above- Wide A-B, far enough out from the PA that there is perceivable cross-signal direct-sound, which is to say, direct sound from the PA on the opposite side is also picked up by the nearside microphone at perceivable level. Potentially different response of the direct-sound cross-signal (the baffled stereo response like thing, dependent on PA coverage angle) with largely different level and arrival time (the stereo hass delay thing).
Fun to think about. As I've posted about numerous times, concert taping of PA amplified concerts is an anomaly in the recording world, partly because the PA is a widely distributed yet highly correlated source, something that does not exist elsewhere in nature nor in the world of musical instruments. That is the case for a typical recording position out in the audience. The close stack-tape stereo A-B (#2) and not so close stack-tape stereo A-B (#3) scenarios are even more unique and unusual. Interesting to think about what is going on and how to use that to our advantage.
[/speculation]