As far as I know, the basic problem is that we don't know whether the stress drop in earthquakes is 5% or 100% of the total shear stress. Certainly it is only a tiny fraction of the lithostatic stress, on average.

Experiments in the lab on cold, dry rock are much closer to 5%. So if the stress drop releases only a tiny bit of the stress, having another earthquake soon is conceivable. If the stress drop is complete, having another earthquake quickly on that same section of fault is less likely.

Two more complications: (1) The stress drop is probably highly variable across the fault in any given earthquake, so some parts might be more likely to re-break than others. (2) If the middle of a previous rupture were to re-break, it might be very difficult to tell whether that particular patch had actually broken in the first event or not.

A messy problem.