I dont believe it would be possible to do what you describe DIRECTLY.
If the aerofoil were free to rotate about the (estimated) AC/NP, the first problem would be that it WOULD rotate - under its own weight, if nothing else. So it'd be quite an effort to provide a balanced aerofoil.
The second problem is that you vary angle of attack in a wind tunnel by moving the model, not the wind. So you actually need to have control over the position of the aerofoil, rather than leaving it free to rotate.
The closest you could do would be to constrain the position though a pivot at the AC and measure the resulting moment on the pivot (perhaps by some simple force gauge) and compare wind on and wind off (which, very simplistically, is what a "real" wind tunnel does when accounting for what's called tares).
It might not be quite as compelling but it'd convince anyone it was proerly explained to i expect.