Howdy Peter......Nicely drawn schematic of the regulator circuitry less the heavy current feed to the 2 pass transistors and the rectifiers back etc. Your theory sounds good....is that what you found in actual testing? I'd troubleshoot something like that by disconnecting the load, disconnecting the crowbar (if constantly tripping), measuring the reg output voltage - if high, move to the base connections of the pass transistors....at this point using a scope to look for glitching while giving that adjust pot a swipe - if glitching is seen (high - near 24 vdc - spikes seen move progressively towards the pot. Get it stabilized by repair then reconnect the crowbar.....or if nothing found.....troubleshoot the crowbar ;)
Note that if the output AND the pass transistor base drive are all high, disconnect and check those pass transistors for shorting. You probably do something similar.
Gord