[Wavytone]

Actually I'm sure they do, but it's complicated and only evident if you do some ray tracing of a Newtonian or refractor, and the eyepiece. The usual criterion used is the diameter of the "circle of confusion" containing 90% of the incoming energy.

Basically, while the curvature of a refractor is of the opposite sign, most serious refractors are f/7-f/9. Most dobs are around f/4-f/5. What this means is that an eyepiece design with a curved focal plane with a curvature (1/R) half that of the dob focal plane produces about the same size circle of confusion in both a dob and a refractor - the net aberration is about the same in magnitude, but of opposite sign.

Similarly, building in some negative coma in the design produces an eyepiece that performs exceptionally well in a fast Newtonian, yet is still quite acceptable in refractors, maks and SCT's.

You could design an eyepiece fully corrected for coma and curvature, resulting in an even smaller circle of confusion over a wide field for a fast Newtonian - as someone did (the Klee "Pretoria" eyepieces) with really superb results, however if used in a refractor or SCT the result will be very poor indeed.

This unfortunately was the undoing of the Pretoria eyepiece as many buyers at the time simply didn't understand it was totally unsuited to their Celestrons and achro refractors. The result was it got a bad name.

Another eyepiece that performed very badly in sone types of scope was the Koenig - though I never figured out what sort of scope it was supposed to match. Field curvature was always a huge problem with these ep's.