So I was thinking something along the lines of incident light readings at various distances from the light source. Taking readings a few degrees off center would be good for beam spread also. We could also do reflected spot readings off of an 18% grey card from the rider's position and move the card to the same points.
Might be more trouble than it's worth.