Are you seriously suggesting to try to back out the external wing loading from FEM type measurements within the structure? That would require a degree of accuracy that simply does not exist, and may not feasibly ever exist in any cost-effective sense.
First, a piezo-resistive matrix on the wing surfaces would give data that eventually could be used to make flying in strong turbulence experienced as rather comfortable.
There would be a combination of inputs needed, with advanced realtime computing involved, surely. A carbon fiber airframe would have more flexing / movement that would make the measurements more 'visible' to the various sensors.
The main 3D physical motion of a wing could be measured from one strategic single point, affordably.