The essential thing is that the sources need to complement each other to make mixing them together worthwhile, and that complementary aspect can take several different forms. As others have mentioned, one good way to approach it is to have each source contributing something different, which the others sources lack.
The more sources you plan to mix together, the more differentiated they need to be from one another. That's more or less directly contradictory to mixing togther mutiple stereo pairs which were each setup to make a good sounding recording on their own.
It makes sense to mix an audience recording made from further back in the room with a soundboard recording or an on-stage recording because each will contribute something different without stepping on the other’s toes. Potential problems arise when mixing things which are too similar to each other, so a general setup guideline is to try to minimize the similarities between sources which you plan to mix together.
If I knew I would have a good on-stage pair and a good PA or soundboard recording, I'd want to setup the third pair to provide what wasn't already well represented in those other two sources. The things I'd want in that source are the room ambience and audience reaction. To get more of that and less of the PA and stage sound, I might put up a pair of omnis farther back in the room, or a pair of directional mics facing away from the stage out into the room. That source will be more useful specifically because it doesn't have as much on-stage or PA sound in it as it would need if it was going to be used alone.
Similarly if you are putting up multiple microphones on the same stand with a plan to mix them together, maximize the setup differences between each pair. Consider spacing the omnis very widely (as wide as you can, wider than you think you should) and arranging the Gefells in the center as an X/Y pair. Those very different setups are much more likely to complement each other without mixing problems than two near-spaced pairs on the same stand pointing the same direction. Using an X/Y configuration in the center means there are essentially only three different microphone source locations being mixed together from the same stand instead of four, so there are less interactions that can produce unwanted comb-filtering. An ORTF pair in the middle may work too, and would be my choice over X/Y if I was using that pair by itself, but X/Y will be a safer bet when used combination with the wide omnis.
As others mentioned, when mixing start by designating one pair as the primary source. Get that balanced and then slowly bring up the level of the other pair to see if it complements things. Play around with the level balance between the two. Try flipping the polarity of the second pair to determine which way sounds better. Start over again using the other pair as the primary one too see if that works out better.
It quickly gets complicated when adding more and more sources to the mix. Anyone can simply mix a bunch of sources together, but doing a good job of it takes concentrated listening and careful adjustment. Done well it can pay off nicely though. Configuring your microphone setups to minimize potential mixing problems makes that job easier and more likely to be worthwhile.
We can provide general advice such as this, but trying various setups and listening to the results while mixing things yourself is the only way to really confirm what works best for your recordings and your ear.