The stages can move for a seemingly arbitrarily long time, but they still overheat if standing still for about 30 seconds. Have to find a solution for that. I don't know if thermal paste will change much when I've already bolted them tightly together.
Major problem: Both stages have a tendency to switch between running smoothly and getting stuck, depending on exactly how the rod is aligned. I'm guessing the guiding bolt is a little loose, and can change between perfectly aligned with the rod and not quite perfectly aligned.
After cutting off superfluous edges and cutting a notch on the Y stage for the tube tightener screws, I have about 11x16 cm of writeable area. Not a lot, but enough for tests and for a RepRap, should I go that way (requires a plastic extruder, though. First some 2D tests!)
Softwarily, I have refactored the little test program to where I have two functions: move(dir, axis, len) and pauseSeconds(seconds). The move() functions is almost calibrated to move in cm.
For playing around, I am tempted to hook up three three-way switches so each axis can be controlled independently. Or maybe get a mini joystick controller:)