While I've never noticed, or at least never been distracted by, FR changes as I stand up or sit down or whatever, they seem to be measurably real in every speaker measurement I've seen that looks at off axis FR. Have I misinterpreted? Isn't this why side-by-side "sideways" centers are disfavored?
Because you move the potentially distracting FR changes into the horizontal plane, where they may become more noticeable (because people lean side to side, or sit at different places on the couch or whatever, a lot more than they leap up and crouch down during normal listening . . .) I frankly thought this was very basic stuff. Maybe I just said it wrong. Maybe I just misunderstood for a few decades.
Yes, precisely. Mostly the typical horizontal center has it all wrong. They are out-of-control horizontal dispersion modifiers.
Some of this is audibly subtle, though, depending on the spectrum of the sound being used and depending on how bad the speakers off-axis response is. In a highly reflective room, it all gets mushy too, so the off-axis issues are less obvious. We can't really think of controlled dispersion as being radical anyway, we're looking at -6dB down points in a polar response, not -20. Even the so-called "constant directivity" horns are not really perfectly constant.
The whole "center speaker" concept was/is driven by aesthetics, which is a valid concern, but the odd/special center is a universal problem not easily solved. Horizontal centers with relatively high crossover frequencies and vertically stacked tweeters seem to fare better.