sounds like a mess.
one of the very basic rules on DCC is to deliver the current to the rails, and test your breakers by shorting the rails.
There’s definitely something wrong with the story, probably someone is not understanding something right.
Poor power delivery to a module could easily make it so a dead short on the module might not draw enough current to trip the main “breaker” that was “far away”. This is typical in many layouts where people do not use large enough gauge wire, poor connections, wrong settings on the breaker.
But what is great about autoreversers is that they are a 4 terminal device. Connect the “input” to 2 rails near the “reverse loop”, and then connect the “output” to 2 rails In the “reverse loop”.
thus a short circuit is really independent of any other voltage drop to the module, or anything.
since these are modules, it would be really, well, stupid not to put the autoreverser right on that module, with leads no longer than about 6 inches.
done this way, there should be no issue at all. I’d bet hundred bucks that the person that got this messed up just does not “see” the wiring or tried to remote the autoreverser, or set the trip current greater than what could be delivered to the module in the first place.
(typical situation, leave autoreverser to trip at 8 amps, but you cannot deliver 8 amps to the input of the autoreverser, therefore it never trips)
Modern autoreversers have a settable trip current. I use PSX-AR units and they are very reliable, in G and Z scales.
Greg