A lot of crazy just happened so I wrote it all down as it happened:
Changing the solving method from 'internal solver' to 'local astrometry' seems to have made it plate solve finally but very slowly
I put 'use scale' and 'use position' on to see if that would help
I started solving again and I caught sight of an error writing log file (screenshot attached) which caused kstars to lock up. I guess this must have been what happened last time too?
I managed to reboot the pi and, when it started back up again, I went straight to plate solve and it died again in the same way. I then went outside to see if the sky was clear and the mount was driving the telescope into the side of the observatory for some unknown reason.
I reset the park position after the strike then slewed to my target (IC63) before plate solving yet again. It appeared to plate solve like it normally would BUT when it slewed to correct, it fired the scope to a crazy position nowhere near the target.
I parked the scope, unparked it, slewed to IC63 and it pointed to a different random part of the sky. I decided that it was not going to recover so powered everything down to start again.
After a power cycle, I unparked it, slewed to IC63 (correctly), plate solved again then it fired off towards the floor again. I checked the sky map and it knew that it was pointing at a random place.
I parked the mount, unparked, slewed to IC63, it went to another random position pointing at the observatory.
I then set the park position to default again, clicked write data, it moved the scope to the correct position but the star map showed it in a completely different place (third picture). I closed kstars, reopened it and it moved to the right place (4th pic)?! There's a bug somewhere, I don't know know what but it's causing me some trouble!!