I think a good deal of Trevors improvements in image quality are due to the new oversize aluminum tube replacing his old cardboard `heat sink '. Increases in flatness on a flat have much less effect on improving Strehl ratio than making similar improvement in spherical aberration on the primary mirror. I once saw a simulation of 1/2 wave P-V curvature on the secondary on an otherwise perfect 16" Newt and the Strehl dropped to only about 0.94. On the other hand if your flat is only really half a wave then theres a 6% gain you can potentially get by replacing it.
I've tested loads of diagonals and particularly in the larger sizes they would average 1/4 wave. I think the high specced ones from Antares are good value. Even a 1/12 wave flat that was genuinely so would enable you to pick up almost as much advantage as was to be had with a good secondary. 1/30 wave secondaries are for the purists only- and I guess planetary imagers are amongst the purists. Remember that an optical system can only be as good as its weakest link
Perhaps I will get to test Trevors old secondary sometime and the reasons will be clearer .
