As you know I am not using the same device as you, but I do have some of the challenges you do.


I don't use an out and back exclusively. I do look at the out and back (if present) as one parameter. Trying to net 0 would work well if wind are "well behaved". But they are rarely well behaved around here. If I tried to net 0 every out and back I'd get quite varying numbers (depending on the weather conditions) and it would insert far more error and variance in runs.

I personally would keep the calibration constant for the entire session. If you have the ability to analyze and adjust post ride, I would do that. I think the ability to analyze/adjust is still in the product. It was way back when.

I find my device is not nearly as sensitive to position change as people imagine. Yes, if you go from just under your hands to the left brake, there will be a huge difference. If you plan huge cockpit changes, adjust the test accordingly. But a new factor for each test will lead to errors IMO.