I have encountered this a couple of times. It is just SCS attitude rate damping, I think. Manual attitude switches in rate cmd, attitude deadband and rate switches in min and low. So the SCS wants to enforce a 0.2°/s rate limit. Now, you start the CM RCS check with no rates, but the way the thrusters are positioned is, pitch thruster points almost exactly through the CSM CG and causes no pitch rate. But the two roll jets you are testing both cause a small pitch rate in the same direction. And if you fire them long enough you get more than 0.2°/s and the SCS tries to get the pitch rate down. But it can't do that as the thruster used for that is ineffective for the CSM. There is also a little bit of a roll that gets induced by the pitch thruster, that is what the roll thrusters try to counter once the pitch thruster fires constantly. I think this is what is going on.
Later entry checklists (Apollo 13+, not that we have much before that...) explicitely switch to min imp mode before the test, probably to prevent this sort of issue. The only thing I could see that might be wrong in NASSP is the exact CM RCS thruster directions. They are not as perfect yet as the SM RCS thrusters. So maybe this wasn't as much of a problem in the real spacecraft. But as far as I can tell, there is nothing in the SCS behavior that changes when you switch to CM control. It tries to use the CM thrusters the same way with the full CSM as with the CM alone, same deadband and rates and all. So I think this is almost certainly correct SCS behavior. I do find it a bit strange though that there is no warning or at least a note in the AOH Volume 2 about this.
