I've seen several posters coming up with similar theories to justify that camera square. None of those theories seemed to have any verifiable science behind them. Not any that I can find anyway. Where are you guys getting these theories? To be fair, you may be 100% right. It just doesn't seem to be valid because multiple OEM's are doing computational photography without that lens configuration. Heck Apple is doing computational photography without that lens configuration.
http://lightfield-forum.com/wordpre.../linx-array-camera-modules-multi-aperture.jpg
When Apple bought Linx in 2015, for their camera array tech, they kinda showed their hand in what future phone camera setups would look like.