I'm sure you are aware there is an active debate whether major and minor faults are strong or weak, based on the rotation of stress approaching faults and around the time of earthquakes, as well as large mountains supported by serious stresses near faults, and heat flow arguments, etc.

The general observation is NOT strong rotation of stress at the time of large earthquakes, so the paper from about 1990 that you cite was premature in claiming complete stress drops in earthquakes like Loma Prieta.