My suggestion to use X/Y in the center is a simple extension of the single forward facing center cardioid (doesn't have to be a cardioid BTW, it could be any pattern at all, it's just that a cardioid or supercardioid works nicely). Actually I'd preferably do it as Mid/Side by just adding a fig-8 to that forward facing center microphone, but X/Y cards or X/Y supers would be fine. That uses all four channels so the taper won't feel like they are 'wasting' an empty channel on the recorder or leaving a mic unused, but still keeps things manageable and simple and allows for some center image spread adjustment if you want- pan them both center and it's as if you had a single forward facing sub-card in the middle, or pan them out a bit or all the way hard left/right for more center image spread. You can pan them like that without problems since they are coincident and won't have phase issues as they mix with each other, yet there is still the good timing and phase difference info provided by the combination with the wide omnis so the recording won't sound flat and lifeless like a pair of x/y cards would alone. At least an X/y pair of cards used alone sounds that way to me.
Here's the technical part- regardless of any panning of the X/Y (or M/S) pair, there are only two phase relationships being mixed electrically- the coincident center pair with left omni and the coincident center pair with right omni. Electrical mixing is where the potential phase problems occur for the most part. There are three phase relationships going on in total including the phase relationship between the two omnis, but the omnis aren't mixing electrically with each other.
A near-spaced pair plus wide omnis can work fine too. In that case, in order to minimize potential phase problems, it will usually be best to just route the near-spaced pair hard left/right so each center microphone mixes only with one omni on the same side. But even doing that, the phase relationship is more complex than having a single microphone or X/Y pair in the center. If direct-routed or hard panned left/right, there will still only be two phase relationships being mixed electrically- the near-spaced left card with left omni and the near-spaced right card with right omni. But there are now not only three but six phase relationships in total. Tony Faulkner leverages that complex phase relationship intentionally, somewhat like the way a phased-array radio antenna works. By very careful spacing and close listening during setup he avoids phase problems. Concert tapers don't have that same ability to carefully monitor and adjust things on setup before recording, so it becomes somewhat more of a gamble. It may work fine though and sound great.
However, if you were to try to pan that near-spaced center pair like I mentioned doing with an X/Y pair, say to fine tune the center image width or maybe solidify the middle or whatever, you are much more likely to get phase interaction problems because you'd then be mixing five of those six phase relationships between microphone positions electrically, instead of mixing just two of three.
Things get complicated quickly when you electrically combine (mix) multiple microphones which aren't placed very far apart from each other. That's the basis for the 3:1 rule for microphone spacing when mic'ing individual sources on stage that will be mixed together. The 3:1 rule doesn't apply to stereo recording microphone setups because stereo uses those phase differences intentionally and the two stereo channels are kept separate and aren't electrically mixed with each other, but when more than two microphones are used in a stereo recording array, at least some of those microphones will be mixed together electrically, and the same phase interaction issues that the 3:1 rule is intended to minimize will be in play.