Well, I kind of rhetorically cheated a bit there with the bit about liking ORTF more than X/Y because of it's phase interactions (talking main pair/audience recording, not close-up, spot stuff). That's true in most cases, but when listening in stereo, not necessarily mixed to mono.
[As an only semi-relevant aside, I also prefer spaced surround configurations rather than coincident ones such as dual M/S or ambisonics for the same reasons. In stereo, the exception is Blumlien, and crossed super/hypers with sufficient angle between them, so M/S can satisfy that requirement. But I haven't found a similar exception for coincident surround recording, I speculate that 1st order patterns just aren't tight enough to allow for a coincident arrangement to work for more than a minimal number of playback speakers, like 3 or maybe 4 max. I'd love to try a coincident array of exotic mics which can exceed 1st order directivity, like 5 Schoeps SuperCMIT shotguns, or the Trinnov SRP.]
As for repetition, below is what I was typing while Jon & D were posting.. enough repetition from me for the next week! Out of town and off line till next Tuesday.
Regarding the time delay causing said phase interference, I think it's worth noting that in the mic placements I'm describing the flanking omni mics are obviously not coincident but are still somewhat close - in the experimental one I've posted here and some of the other spacings linked, each omni mic is 10cm from the inner subcardiod. I know that is close to the wavelength of 3400 Hz, but I'm not sure if that means that's the frequency above which there would be phase interference between each "side" pair.
The "still somewhat close" region is where the dangers lie. Neither fully coincident or far enough apart to avoid (theoretically) or sufficiently minimize (practically) the interaction problems. That's the "tread carefully, and it's good, or suffer problems" zone. At least for the midrange and above.
I'm more a spatial geometry guy than a numbers guy, but let me see if I can get this right-
Things get interesting at half that frequency. Lets say one 3400Hz wavelengh cycle = 10cm (not exactly, but for purpose of discussion). For a sound of that frequency arriving from directly in front, perpendicular to the array, the arrival time is identical at both microphones and you get in-phase summing (which is the case for all frequencies, not just 3.4k). You also get in-phase summing if the 3.4k source is directly off to either side, 90-degrees to your mic array, upon completion of one cycle. In that case the sound arrives at the second microphone delayed by exactly one cycle. But at an angle somewhere in between, and here's were my trigonometry gets rusty, the path length difference will be exactly 5cm, the resulting microphone signals will be 180-degrees out of phase, and you get a cancellation notch. That notch is symmetrical to the pair however (the same for sound arriving from 90 degrees left or 90 degrees right) so you get one cancellation notch pointing left and a mirror image of it pointing right.. Actually that's a simplification to flat plane, in reality it's 3-dimensional and forms the surface of a cone pointing left and and identical cone pointing right. For the angles of arrival between those positions you get a smoothly varying gradient shifting between cancellation and reinforcement.
At 1700Hz you still get full summation from directly ahead (no time/phase delay), but from 90 degrees left/right you get 180-degree phase cancellation. If you plot all angles of arrival for the 1.7k sound, you get a nice, neat, forward facing figure-8 response, but only at that single frequency.
Below that frequency you still get forward reinforcement directivity, but without a deep notch directly to the sides. The response broadens out until it eventually becomes omni-directional.
At frequencies higher than 3.4kHz a 10cm path length begins to span multiple wavelengths, so you get more than one notch and multiple lobes, until you end up with a fan-shaped array of lobes and notches by the time you get up to 15kHz and beyond.