About a zillion :-) Building one has been one of my "wouldn't it be neat to do ..." projects for a long time.

Can you say a bit more about the lenses? The Heathkit oscilloscope I built as a kid had electrostatic plates that were used to focus the beam rather than anything lens like.

Have you considered using bits of lead type as your test target, they have the advantage of being asymmetric and recognizable, could help with linearity measurements.

The SEM I used at Intel a couple of times had the electrons coming in at an angle and the detector at an angle. have you considered something like that to improve the signal pickup of your optical tube?