Based on recent MTHexapod testing it appears that the STOP command is sometimes ignored. The CSC runs the following command sequence:
If the pause is too short the stop command is ignored and the move finishes normally.
I consider this a serious, high priority bug that requires a lot of tricky code to work around in the CSC – code that delays sending the STOP command.
Please also check if the rotator has the same problem. If it does then please file a ticket or ask me to do so.