Originally Posted by rpenmanparker

A data smoothing algorithm would be advantageous.

And there would be a similar problem as we assumed for the Arofly, that the system would have to be calibrated for the wheel/tire combination. Angular displacement will mean different power input for a stiff wheel than for a flexy one. In other words the same hub windup (what you are measuring) will not mean the same power in every wheel. How would you calibrate the device for the specific wheels.

Originally Posted by rpenmanparker

Power Taps sample at the rear hub without a problem, but obviously more frequently than once per wheel revolution...I think. But I don't think it is important to have the instantaneous power readinf associated with every pedal stroke or showing power as you progress around the pedal circle. A smoothed, rolling average of data taken ever second or so would be fine.