I just finished manually measuring the hub windup on my bike for various weights. Basically I strapped a laser pointer to the tire, kept the crank from turning and hung known weights from the rim, and marked the spot on the wall. Everything horizontal to keep things simple. I used 5 pounds, 10 pounds and .3 pounds.

There was about 1.5% variance in measurement, a little movement of the crank (some fraction of a mm) and I did not account mathematically for the movement of the pointer, nor the weight attachment point for that matter, but I doubt that it's more than .5% error from slop and I'm not looking for exacting precision at this point.

There is a good bit of wind-up, more than I expected in fact, and it behaves linearly as expected so that's all encouraging. Force in kilograms equals 7.47 times the radian measure of the rim rotation. Or, a point on my tire moves approx .35°, or 2mm, for each pound hanging on it.