Love the project, working on something similar myself. Is the software available? I have been struggling to find anything like it. 

Interesting projects and adopted solutions. I am also thinking to the design of something similar using a Proxxon small lathe. I have not yet considered the software aspect of the motors controls, but sure using a standard gcode is a great solution. I plan to start this project around December. In the meantime, I follow your progress.

Sorry for the late reply.

Yes, smaller motor is because of space constraints - I want to be able to close the cover. There is room for a slightly larger motor but not sure that is needed.

The new leadscrew has a pitch of 4mm compared to 1.5mm for the original. What I have found is that the new one tolerates significantly higher acceleration and speed without stalling, at least 4x settings for the old one. Old one had 1.5mm of backlash, new one is a second hand C2 grade screw so the remaning backlash (0.02 - 0.03mm?) is IMO mostly due to tolerances in my homemade cross slide. A huge improvement!

The last few days I have been struggling with spindle sync for grblHAL, seemingly due to too many steps beeing issued. However, it was due to the power supply randomly current limiting during the cutting phase. Strangely it did not occur during rapids. I am just back from the workshop after fixing this, and now multiple G33 cuts lines up perfectly as far as I can tell.

It's amazing that you can fit so many components into the original driving belt and gears train cover. I can see you have a smaller stepper motor? because of space constrain. Do you think it might be an issue if you replace the leadscrew with a ballscrew with different pitch? I don't know,  I just guess because my stepper runs a radio of 2.5 to 1 for the Z axis using the original pulleys,  btw I have the Compact 5 PC, I still lost steps sometimes.