In the theory, as in many theories of instabilities, the size of the disturbance that goes unstable is probably far smaller than the resulting ground breakage in the earthquake. So the volume over which precursory deformation might occur could be as small as a cubic meter, effectively completely unobservable, and such disturbances could be happening all the time with most NOT leading to an earthquake.

Bottom line is that it is perfectly in accord with the proven observations for there to be no way to spot the precursory deformation indicated by the rate-and-state theory.