Glen...I had been thinking that I like your approach....I think that looking carefully/scientifically at patterns has some chance of success, although the success may be rare. As for rates of extension vs rates of strike-slip: the full extension rate across the ocean spreading centers in the Gulf of California should be equal to the right-lateral slip rate on the transform faults that connect them. But, the crust is thin and new at spreading centers so extensional quakes there might be numerous during eruptions, but probably not large. I know there are large normal faults in Baja, some are inactive, some active: I imagine that the rates would be (much) lower than the 3+cm/yr (4+?)on the transforms in the Gulf. One place to consider is the Brawley zone south of Salton Sea, which is essentially a partly formed, obliquely oriented spreading center. The oblique slip may be partitoned between strike-slip zones and normal faulting zones. It looks like from what I have heard that the San Andreas fault itself continues into at least the northern Brawley zone. Armburster and Seeber located 1000s of EQs with focal mechanisms and these are probably available on the scecdc (data center) site somewhere. Magistrale may have done the same (probably by different method) more recently, and these probably also are available. There are 100s of thousands of relocated quakes in the SCEC Community Fault model, but they probably do not have slip (focal mechanisms) as part of what was loaded. They have some free softwares to view the community fault model....but might be time consuming to figure all this out...see link:

You may want to look into the controvery about accelerating moment release: David Bowman and others have been publishing on this, and presented it as a speaker at SCEC annual meeting a few years ago. But, there was a poster at the SCEC meeting saying it was not correct (I could dig up the reference to that poster on request.
Chris

• harvard/SCEC CFM