The alignment plan builds a pointing model in EQMOD. Essentially a single alignment point behaved like a single star alignment using the hand controller, with pointing accuracy falling off the further away from the sync point you are. Two sync points it chooses the nearest one and offsets from that. Three or more, if the target falls within the triangle made by three points is very much more accurate, if it is outside it goes back to the nearest point method. It allowed me to build up a web of align points so that anywhere I wanted to point falls between them (And they are ALT-AZ points not RA-DEC so they don't move with the sky) and it reliably landed anything I asked for almost perfectly in the middle of the chip.
Possibly better value in visual than imaging but I have had a tendency since starting the imaging to hunt around for targets during the night, looking for things that suit my framing to image properly later so it is handy to have what you asked for land right on the chip every time.
Plus it is cool watching it do its own thing aligning itself
