Brian,

Yup, Atwater 1970 first interpreted the whole San Andreas fault as a transform linking the Mendocino transform-transform-trench triple junction to the Rivero triple junction in the south....I think it is a transform-ridge-trench junction (?).

Actually, transform faults between active spreading centers in the deep oceans are very active. The slip rate is equal to the full spreading rate. This can be as fast as the San Andreas...or faster. But, they are only active between the spreading centers: their fossil traces in the ocean crust are called fracture zones.

It is pretty common to have M6 quakes on them. But, M7s would be rare, both because many of the transforms are short (relatively) and near the spreading centers the brittle crust is very thin and weak...or maybe I should be talking about the lithosphere...same deal.

There are a couple of really long transforms at big offsets in the plate boundary...for example, in the Atlantic near the equator. There were a couple of larger quakes there in the last couple of years.
Chris