Thanks.
One thing I've read is that many people see a small amount of pincushion (up to 7%) as undistorted, which agrees with your findings.
If the eyepiece had little RD and had, instead, AMD, the effect could be additive, making AMD appear worse.
I only have one eyepiece my eye perceives as having a completely flat field, and the rest vary but mostly appear to have a slightly concave appearance, just like the night sky when you look up and see the entire sky.
(The night sky appears flat at the zenith but seems to curve down to the horizon--call it a large, flattened bowl) --flat in the center, then slightly curved upward to meet the field stop.
One thing: a curved focal plane will be perceived to be flat so long as the focal plane curvature of scope and eyepiece are identical. I suspect that the curvature may influence one's perspective,
but if the stars are sharp to the edge, how would we tell if the field is flat or curved? Probably only by panning. To me, that means that strong RD or strong AMD are negatives when panning the scope.
But, if neither RD nor AMD is solved for in the design, how much of each should be left in? I don't have an answer.
One of the perceptions I haven't explained yet is that the field appears to be different distances from the eye in different eyepieces.
I once thought it was the apparent field, with wider apparent fields appearing nearer the eye, but then I discovered different 100° eyepieces appear to have the field at differing distances.
So, I cannot explain why different eyepieces appear to have their fields at different distances from the eye. It could be the influence of eye relief, but I am not sure of that.
I am so glad not to be an eyepiece designer. It's obvious to me some factors are often ignored, like EOFB, or the focal plane distance perception, or how to design an eyepiece to appear as if it has the least amount of distortion for the eye.
It seems a designer would first have to become an expert in human vision.