I'd just like to confirm that X/Y = "coincident" as opposed to A/B ("spaced"); it's as simple as that. All that's implied is that the two microphones will have the same directional pattern as each other. There's no implication of what that pattern is, or the angle between mikes.
If you travel along the spectrum of first-order directional patterns--from omni at one end through cardioid in the middle to figure-8 at the other end--the farther you go toward figure-8, the narrower the front pickup pattern becomes. And the narrower the pickup pattern, the narrower the angle between mikes needs to be so that you get smooth, continuous coverage in the combined stereo pickup.
I think pretty much everyone knows that for figure-8 microphones, the angle that gives you that smooth coverage is 90° (the so-called "Blumlein" arrangement). That's a good point of reference to keep in mind. By comparison, if you take microphones with wider front pickup patterns and set a 90° angle between them, you'll get a more and more center-heavy pickup. The two channels will have more and more material in common between them, i.e. the recording will be more nearly mono, and correspondingly less spacious, less qualitatively responsive to the room's characteristics, AND less specific as to localization of individual sound sources.
Even with supercardioid microphones you can go well beyond 90° and still have smooth, continuous coverage, thereby getting a much more interesting, positionally definite, and spacious-sounding stereo recording. The idea of using 90° for X/Y cardioids as a "cookbook recipe" is just sad in my opinion; a much better starting point would be 120°, and in many situations one might well go beyond that angle, depending on the room, the recording distance and the particular microphones.
It's especially important, I think, to use loudspeaker playback as the basis for deciding this, rather than headphones, unless you're recording for the purpose of headphone listening specifically. With practice you can learn to correlate the two somewhat, and make decisions via headphones as to what angle between microphones might sound best over loudspeakers--though it's never the same as what sounds best through the headphones themselves, and I still get this wrong occasionally.
--The one virtue of X/Y cardioids at 90° is that they pick up a very wide angle directly--essentially the entire front hemisphere. When your microphones are only a few feet from multiple or very wide sound sources, that can be a good thing. Recorders such as the Zoom units, or the Røde and Sony stereo mikes that have X/Y cardioids built in at 90°, are sold largely to people who want to record meetings or group discussions at relatively close range. Plop the recorder or the microphone down on the desk or table that people are gathered around, and it will pick up 180° nicely, with significant pickup still beyond that angle.
When you're a music taper and you mostly hang out with other music tapers, it's easy to forget that we're only one sliver of the market, and that the equipment available to us is often manufactured for speech pickup applications that have different requirements from ours. I've been on both sides of this, and any time I've written marketing copy where I could say, "This microphone has often been used successfully for ... " without lying absolutely, I've done so, or at least felt the pull to do so. But as someone who buys and uses this equipment, I try to be more critical of how things are marketed, and not rely too heavily on anyone else's conscience.
So if you keep seeing equipment with cardioids fixed at 90°, don't let that image seep too deeply into your visual cortex; use your ears instead.
--best regards