I'm a little confused about your approach and perhaps understanding it will help you too.
Plate solving its self is just the process of matching an image to a database. What you seem to be after is plate solving together with re-slewing to original co-ordinates.
You seem to be trying to give the system an image from last night and match that location. I think instead you need to:
1) go-to desired co-ordinates
2) plate-solve and re-slew so that your telescope is now truely pointing at those desired co-ordinates.
3) take images.
4) packup.
5) start a new night.
6) repeat steps 1, 2 & 3.
Point is that you are slewing to co-ordiantes not to match an image, but you are slewing to co-ordinates with certainty so hence your images from multiple nights end up of exactly the same field.
The above can be achieved by using the software you have together with CCDCommander which is relatively inexpensive, but does still cost something. I'm not sure if it can be achieved without that, by using just Maxim + TheSky, but I don't think so
