I'm just curious as you trace the bonding, where does the bonding system connect to the DC main negative side of the batteries? If it has a good resistance free connection to DC-, this must have been a short that preferred going into your bonding system and perhaps finding ground through the water(your corrosion on the "terminals" underwater) to engine via stern tube?

I think my bonding wires that are connected to chain plates, through hulls, etc, use the large plate steel fuel tank tops as a buss. From the tank tops, the bonding system connects to the DC Neg via a wire to the engine block. But I'll be honest, I have to check to be sure. Maybe this was not done on my boat? Your scenario will have me checking!

For instance, an original 51year old wire(which I have lots of), running past a chain plate(behind the wood ceilings), could chafe into the + wire. If my bonding system is not connected to the engine as I suspect it is, it would not likely short, and blow a fuse(20 amp on the 12 gauge line).

But anytime the lighting circuit was energized, this could energize all my bonded-non current carrying terminals, including through hulls. Then the stray current may find the nearest ground through the prop shaft or bronze stern tube which are both connected via the engine to the DC neg side of my batteries.

Your scenario just has me thinking I should just check to be sure my bonding system also provides a separate path to ground in the event of a short circuit on anything connected to my bonding system. Thanks for bringing this subject up.