I'm a latecomer to this thread and I haven't read the preceding posts, which may doom what I have to say to irrelevance, but: A third, center microphone in a two-channel stereo recording is useful when there is something particular for it to pick up--something that is available mainly to it, and not (or not clearly enough) to the other two microphones. Decca and RCA classical records used three main microphones (and three recording channels on 1/2" tape) for many years, with or without acoustic baffles to increase the separation of sound pickup among the three microphones at mid and high frequencies. (Decca went through far more variations than the public seems to know about; there's no one cookbook formula for a "Decca Tree" that deserves any special prominence over all others.)
But if the center mike is only going to pick up another "spatial sample" of the same general sound that the left and right microphones are immersed in, then it has a rather different kind of effect, which is to scramble the phase relationships in the recording and produce "comb-filter" effects--which can be interesting and useful, or harmful, or meh. To a great extent it's a game of chance. Sometimes it helps if you filter out the bass from the center microphone, and sometimes not.
Just make sure that you record the mid mike on its own track, because when you get home and review the recording, if you're honest with yourself, half the time or more you will find that the input from the center mike changes the sound but doesn't really improve it. There's something about listening to the output of two microphones with two speakers and two ears that is more engaging and less tiring than listening to multiple microphones working out their petty conflicts. Try not to be biased by the expense and trouble that you've gone to in setting three mikes up rather than two, if you can.
--best regards
P.S. added later: I see that one issue is the attempt to get more bass when the main microphones are cardioids. Proximity effect can't be used for this in most cases. The situation can definitely be helped by taking the low-frequency output from one or a (spaced) pair of pressure transducers. But mono bass is inherently boring, so the spaced pair is enormously preferable in terms of the spaciousness of the result, and four tracks rather than three are then called for. As a crossover frequency I suggest something quite low such as 50 Hz.