First, what exactly do you mean by “get the phase as close to +1 as possible”? Are you proposing to move the clips so that their temporal alignment is changed? I mean, I’ve done a lot of that in the past, and it has always ended up in tears.
So in answer to your question, I would do none of the above. I would mix it and get it sounding as good as I can. If it’s obvious that some sort of phasing/comb filtering is screwing up the overall mix, then obviously you will need to check it out, see exactly what the problem is, and what can be done about it. Every case is individual, I don’t think you can apply systematic rules and procedures - and in any case, if the only problem you’ve got in the mix is a bit of phasing then it’s not really a problem. In any case, phasing can enhance in some circumstances - it’s not like the existence of phasing is a blight on all mixes, it’s only a problem if it doesn’t sound good.
Can you clarify your criteria for selecting/de-selecting specific mics?
Next, I don’t understand this:
Why do you want to mix differently to how it appears just for TV?
But more importantly for an audio point of view, how are you going to create such a contrived stereo image when all you have is mics suspended from a ceiling?
From a technical point of view I’d be choosing the best-sounding stereo pair and then take it from there. I’d like to bet you could get 90% of what is required from a decent-sounding pair. That would probably help with mono playback too.