I think your view is in line with many of the experts, and the rest are split between believing characteristic earthquakes/seismic gap theory more and believing it less.

Some fault segments may indeed act characteristically, or maybe none do, or maybe faults act repetitively for a while then switch modes. A topic of much active research.

Given that it is hard to go out to measure a thousand years of seismicity to test the models, much effort now is directed at probing the relation between fault structure and earthquakes, in the hope of foretelling at least the segmentation of the faults and perhaps the state of ripeness from such mesaurements.