Originally Posted by gigaplex

I don't think you can draw that conclusion. The settings R1 use are track tested, and all the other settings you've checked in that table are quite low for a stock motor. There's no way they'd be using that setting if it was obviously that bad. Pretty much everyone who uses an R1 in my area uses more than 45 endbell timing.

I'd expect a faster spinning motor to have a bigger KV drop simply because a slave motor produces more braking torque the faster it spins. And since power = torque x RPM, and braking torque is a function of RPM here, you end up with a quadratic function of RPM vs power in these tests. But these tests don't gauge peak power of the test motor, they gauge power of the test motor at the point where the decreasing torque output becomes insufficient to overcome the increasing torque of the slave motor. Peak power is at approximately half the no-load RPM, so I'd expect a ~50% drop in loaded KV if it's really at peak power output. These tests are well below that range.

If we ignore all that and look at it another way, look at the loaded KV vs timing. For each 5 degrees of timing you're adding, you're gaining approximately 100KV of loaded timing. There's virtually no fall off on that specific metric in these tests. Adding timing is still making it spin faster even under a load.