Yes. I was looking at the gain of the response at low frequency and noticed that it was not close to 1. That is why I had you increase it.
I don’t know if you have looked at your logs but there is a lot of missing data in your logs. I don’t know what causes that. you can try deselecting FAST ATTITUDE in the log_bitmask. The other thing that concerns me is the oscillations at the rotor speed that are seen in both the pitch and roll signals. I saw that you are using the harmonic notch but the vibration is still affecting the signals.
The reason the roll is not successful is due to it not finding the max allowable rate P and rate D. Try to run this test one more time and see if it is able to complete successfully.