Might be.. maybe in its interaction with the signals from the mics spaced away from it. It should in theory retain the the same phase relationship with any mics it was coincident with regardless of pan position, excepting any artifacts from a real world realized, never quite totally perfect coincident alignment.
How well left/right balanced was the sound in the recording position to begin with during the performance? Was it well balanced enough that the decision to point the shotgun toward one PA rather than the other was arbitrary or was that side significantly louder?
One thing I like about these arrays is a somewhat greater ability to re-balance a lopsided image afterward. Usually I do that through some combination of level balancing the various pairs in combination with panning the center Mid - whatever works and sounds right. Can only do so much with the spaced pairs level-panning wise without throwing the overall level/power distribution out of whack, so panning any mono center contributor becomes a powerful tool for centering things. Even though that also alters the power distribution it seems to do so less severely for the same amount of correction in the mix. If I get the power distribution of the spaced mics about right, I can sort of force center balance with center signal panning.
The thing is, to bring the center image back to where it should be by panning the mono center stuff, I'm compensating by panning toward the quieter side. What's odd is that's the opposite of what you describe (I'm presuming here that the center shotgun was pointed toward the louder PA side), and yet the resulting playback image remains balanced with it panned the opposite way, toward what was presumably the louder side rather than away from it.
Part of what may be going on is another thing I like about the arrays subjectively, yet sometimes nags me a bit philosophically in an objective sense, which is this- L/R imbalances seem less problematic overall and somewhat less nakedly obvious in general. Sure I hear them and work to correct them, but they often aren't as severe as I expected them to be when I was making the recording. Now that could be that in order to listen to the recording I'm going through some of that balancing process anyway in getting the various channels roughed in (less the center panning part). What nags me is a niggling thought that if it were unbalanced live, I'd think that imbalance would be conveyed in something of a more obvious (accurate?) way, even though to some extent it is, just somewhat less than I expected. Might be that the well-conveyed, general sense of openness and immersion, which tends not to be unbalanced, sort of covers for that.
Then I remind myself that this is all an auditory illusion anyway, and the ultimate goal of these arrays was never absolute localisation accuracy which perfectly corresponds 1:1 with real-world source placement, but a convincingly real sounding recording with seemingly accurate and precise source placement that supports the musical listening experience. Maybe these arrays are sort of applying the same spatial filter to everything, but I try hard to listen for that and am not able to identify it in any obvious, or at least negative, way.