We have discussed IMU here:
Whether you can get pitch I hope I can try. But the reference also says lean angle is hard. I would have said it is worth a try. I have not tried on the bike yet. But I have made some tests and also put a few graphs in that thread.
Shaking two different IMU cards didn't upset the gyroscopes so much. But it is just "a hobbyist's simple test".
I hope I can continue experimenting. At least it is very interesting. And IMU:s develop fast, partly due to drones developing.
May have to leave the ultrasound a while. Ran into what I believe is "blocking code". I added some very simple code and the program suddenly went "bananas" 
But I hope I can attach the ultrasound sensor to the bike again and see if results differ when the bike is rolling and fresh air flows over it.
I use ESP32 for the IMU. Is that less likely to suffer these "strikes of lightning" of "blocking code" that sometimes occur? ESP32 has more memory than Arduino.