I'm now pretty convinced this effect is real. I'd assume it happens in any porro and becomes visible in ultra-wides because they're pushing the limits of the optics. I'd surmise that prisms, not lenses, are the cause, and it likely has to do with the sequence of the prisms. I've believed that one prism handles the vertical flip and the other handles the horizontal flip. The prisms in the porros I've just looked at are mounted at right angles to each other but "diagonally" with respect to the horizon, however, so it's not immediately obvious which one (or if only one) handles the lateral flip. Is it correct to surmise that the first (back) prism is the one closer to its limits in an ultra-wide, i.e., under more "optical stress," and the more likely cause? I'm a bit embarrassed to ask, since I used to think I knew how the darned things worked.
Edit: D'oh! There's no "vertical" or "horizontal" in the raw image apart from our interpretation of it. When you look at something and tilt your head, the image doesn't tilt. The fact that the prisms are 90* apart properly inverts the image. The mounting angle of the major axis of a prism with respect to the horizon at any particular moment doesn't matter. Thirty seconds with a pair of loose prisms will prove it.
Edited by asphericalaberration, 24 May 2020 - 01:12 PM.