The way I've tested this is to put a big spatula or something to block the arm that you're not testing, and then put big bowls in the lower rack in such a way that the end jets of the arm can still hit the door. Then just run and listen to the rate that the jets sweep the door and the corners of the door. Depending on what parts of the arm have jets at the very end, you may have to divide by that number, for example if the arm has end-jets on opposing ends, you'd divide the rate of the sound by 2. If you heard a swish every second, you'd know that the arm is actually rotating once every 2 seconds, so that'd be around 30 RPM.
100 RPM actually seems like it would be quite fast, especially for the jets to be coming from 4 arms instead of 2. But with the volume and pressure coming out I'm sure it didn't affect performance at all versus the arm moving slower. I've always thought the common speed for an arm to be around 30-45 RPM. Even 60 seems a bit swift.