> 1) Would I be gaining new knowledge, or would I be wasting my time by
> re-inventing the wheel? If there are any internet links to the test that
> was already done, I'd appreciate a post...

I would think that you would be gaining new knowledge because it would come from your own experience.

> 2) What would be the best test to perform? I was thinking of using
> 12" square panels, supported (but not clamped) on 2 edges, with an
> increasing load placed on a board running down the center. (a 3-point
> bending load)

one thing to consider with 3-point bending of a plywood laminate is the orientation of the 2 of 3 plys, just be consistent.

> 3) What would constitute a panel's failure? Audible crack? Revealing of
> cloth weave due to stress? Wood broken through?

take all the data in, amount of bending per load?, first fracture? complete failure? unless your going to make numerous >3? tests of the same layup I wouldn't put too much stock in precise numerical data. I think just seeing how things fail is worthwhile. I would be interested in whether different orientations of multiple layers affects characteristics of failure. How about trying strength of minimal seal coat and fill coats compared to soaking wood and fill coats? go for it