Perhaps for drag, you could measure the torque on the vertical rod used to measure the lift. The guide for the vertical rod could be mounted on a pivot, and the reading from the force cell (either above or below the pivot) could be multiplied by the inverse of the ratio of the distance from the pivot to the airfoil and the pivot to the force cell (I think this is right  just doing it in my head which isn't overly reliable). This would eliminate an extra rod. Just an idea. If the angular displacement at the pivot would be over 1 deg. then it might be necessary to rig a parallelogram linkage to adjust the airfoil angle to compensate.
