I found a discussion in the issues about making it obvious which way things are moving when commands are given and one of the comments mentioned pageup/pagedn which seem to work for focus. Mouse wheel does seem to work as well.
Autofocus doesn’t seem to be doing much but that may be because I just have a piece of lens cleaning tissue on there right now and there are many levels it can focus on. I’ll dig out some prepared slides tonight to try.
But overall glad things are working!
Brilliant! Yes, lens tissue will confuse the autofocus algorithm, which definitely works best on a nice flat sample (as you’d expect). At some point it would be great to do an autofocus based just on one region of the image, by reducing the camera’s region of interest before, and increasing it afterwards - that wouldn’t be particularly tricky to implement, but has never been much of a priority…
Confirmed autofocus works better on flatter samples 
Though it does seem my Z motor may be a bit iffy as it seems to move much easier in one direction over the other - even when not connected to the microscope. But as cheap as these little steppers are I’ve come to expect the occasional dud with them.
I soldered up 3 more boards Sunday evening. 1 is waiting for me to order more components (of course 2 of the capacitors I didn’t order spares of decided to go flying after coming out of the cut tape!) 1 needs a bit of re-work (F1 and D1-4 probably need to be fiddled a bit due to the footprint mismatchs) and 1 appears to be working perfect. So I’ll hopefully have 2 spares in a day or two and one more that will be complete whenever I order components for some other project and can slip a few extra caps into the order.