A couple of things. For what it's worth, just to prevent speculation from becoming established as fact, both the MK 21 and the MK 22 were designed primarily for use as spot mikes in studio recording. Stereo recording with pairs of these capsules -- obviously not coincident pairs, but somewhat closely spaced, like ORTF with cardioids only farther apart -- was truly an afterthought as far as the company was concerned at the time.
The MK 21 was introduced in 1988 (and was followed by the MK 21 H, with a mild high-frequency boost for "pop" recording, about five years later) while the MK 22 is from 2009. I was in the "beta test" program for the MK 22 and was the first person, I believe, to use a pair of them for a stereo live concert recording. Everyone else in the beta, as far as I'm aware, was using them one by one as spot mikes; the capsule's stated purpose was to capture as much of the MK 21's sound quality as possible with near-cardioid directivity, so that spot or solo miking in studios wouldn't pick up other, nearby sound sources or as much room sound.
The special stereo bars for these capsules (STC 21 and STC 22) were introduced only as a result of special orders from, ahem, certain rather persistent customers in the United States, and then when those initial batches were sold, the company decided to make them generally available. In the case of the MK 21 it was 18 years between when the capsule was introduced and a stereo bar became available for it (!). For the MK 22 it didn't take so long, fortunately.
--In my experience, stereo recording with the MK 21 has something in common with Blumlein stereo recording: The results can be near-ideal, but the real-world situations in which those methods are the optimal choice are few and far between. When a pair of MK 21 is in the right place at the right time, beauty happens--no doubt about that. I only wish that I could use mine more often, but I can't. The MK 22 is better adapted to a much wider range of real-world recording situations in my experience. I think I've used my MK 22 maybe five times as often as my MK 21, and I've had the MK 21 for longer.
But keep in mind, please, that (a) I record classical concerts that generally have multiple instruments and/or voices representing discrete source locations across a left-right spectrum, with a front-back depth dimension as well. The goals aren't necessarily the same as when picking up an amplified mix from P.A. loudspeakers that don't, in general, represent a spatial distribution of discrete sound sources that you want to represent in the recording. (b) I record only when asked to do so by a performer or performing organization--so I get to decide, or at least negotiate, how many mikes to use and where to put them. When I have to record from farther away than the sonically ideal location, either some beauty has to be sacrificed for the sake of clarity (hello MK 41 and/or some kind of M/S pair)--or else imaging has to be given up (or even sabotaged), and then more widely-spaced microphones may come into play as possibilities. Again the MK 21 is definitely an interesting candidate for that kind of recording, but that's not a situation that I'm in often.
P.S.: With all due respect, I don't know of any reason to believe that the polar diagrams for these capsules are wrong in any fundamental way. In particular the MK 21 has the most similar frequency response at all angles of sound incidence of any capsule on "the omni side of cardioid" that Schoeps has ever made. That was a basic principle of this capsule design: to have no off-axis peaks like in a typical cardioid, nor an on-axis peak like an omni, but to dwell at the point where the two effects cancel each other out the best.