It's been a while, and I got sidetracked by other things a bit but there's been some progress to report.
The spherical test plate has been finished to reasonable sphere (conservatively 1/10 wave ptv). It does show some TDE but test plate is nearly 20mm larger than a secondary itself so I'm not concerned that much. First image shows a Ronchi snap - smooth and straight. Foucault shows that any zonal differences are very hard to measure, less than 50 microns of LSA in any case.
The convex secondary was then finished to match the test plate. It is still a bit more convex (approx 1/4 wave) but it is a sphere, which is all it counts.
When secondary is placed in near contact against test plate this is a true interference test, so I could use OpenFringe to analyse it. It gave better than 1/30 wave RMS on the wavefront. This is not going to be a planetary scope so I'm happy with that.
The rough appearance of interference lines is an artifact of test apparatus - the 45 deg semi transparent mirror has been made from some sort of foil which is quite uneven optically. When secondary/test plate stack is viewed directly lines are much more smooth, but I can't take pictures of it easily...
The 315mm primary has been now polished with pads, ready for figuring. It again shows a reasonably smooth sphere, as seen in this Ronchigram. All it needs is now a smooth 67% ellipse...
The optical design is now finalised, I could squeeze just about diffraction limited performance over full 35mm chip at rather fast for a Cassegrain f/5.5, with minimal vignetting. Of course, obstruction is nearing 48%, but that's the price I'm quite willing to pay. Most commercial Cassegrain systems have much worse vignetting profile, and users don't complain much
The mechanics are also coming together, I'll snap some pics to share soon.