Chip,
The frames we tested were basically seasoned. EMC in Laramie is about 6%, so things dry out pretty fast here. It might be possible to predict stiffness as it varies with MC, but it wouldn't really be useful for frame response. Actual member stiffness is not really that important. The joints themselves make up the lion's share of the displacement of the frame under lateral load. Natural looseness of the joint, bending and shear flexibility of the peg, and localized crushing of the base material all contribute more to frame deflection than bending of the frame members themselves.
Our objective was not really to evaluate the relative merits of different joinery styles (splines vs. M&T, etc.). We wanted to develop analysis models that can be used to predict response of whatever joinery was used. In general, from a strength perspective, you'd be better off to use splines rather than M&T joinery, assuming that the splines fit into the frame (You don't want them sticking through an exterior wall adjacent to a post). Using splines makes it easier to guarantee that you have adequate spacing and end distance on the pegs and also avoids the problem of tension perp. loading of the post.
You're correct that longer braces are better (stiffer). The question is just how much better can be answered by doing a parameter study using structural analysis models (normally computer programs). Again, the results of our work will help you do that study by giving reasonable values for joint stiffness to plug into the analysis models.
We haven't done any work with scale models. They would certainly require less material, but the time required to cut them, test them, analyze results, etc. would probably not be affected signficiantly. Also, with full-size frames, we avoid questions about the scalability of the results.
Our joint tests (see the Joint Report and the Long-term Report) included SYP. Good stuff!
Thanks for your interest and questions.
Dick