Originally posted by Ed Hurst Of course, if you focus with magnified live view, then you will automatically be adjusting anyway for any differences in refractive index (with a margin of error, that would probably apply anyway).
I was also thinking of possible secondary effects such as degrading the correction of spherical or chromatic aberration, or of flatness of field. Such effects, if present, would be magnified by comparing a scene image with and without the clear filter. I'm personally unsure, however, just what type of scene would be good for that kind of test.
I would expect a thin gel to be pretty minimal in effect, but that too could be tested with a suitable target scene.