I started out the day kind of floundering around. The misbehaving oil pressure alarm which has been joined by a misbehaving glow plug relay had me a little stymied. I disconnected a few connections and cleaned them but of course, that didn't really fix the problem, nor did I expect it to. I was just stalling.
Finally, I studied the drawings and pulled out my multimeter and got busy doing actual logical troubleshooting. It's really nice to have accurate drawings.
I eventually got to the point where a wire that shouldn't be energized was, but not with the full 13 volts but rather with about 10 volts. "Hmmmmmm," says I, "Most curious." I kept tracking and sure enough, I found the culprit.
The engine components are connected to the control panel via two wiring harnesses. To make it easier to install and/or remove the engine, the wiring harnesses are broken in the middle and then connected back together with a multi-pin plug like they use on trailers. One of the plugs has 4 pins and the other has eight.
The battery power was on the terminal with the red wire. Right across from this terminal was one with a black wire which was energized only when the ignition switch was in the ON position. Or, at least, that's how it was supposed to be. When I pulled the connector apart, this is what I found:
Well, actually that's how it looked after I dug a little chunk of what looked like molten copper from between the two sockets. I figure that what happened was that some seawater got between the two halves of the connector and caused a short between the red (hot) socket and the black (switched) socket. The intermittent current flow undoubtedly caused some arcing which results in high temperatures which probably melted some of the copper in the socket and made a more or less permanent, high resistance connection between the two sockets.
My fix is not as elegant as the 8-pin connector, but I was able to use stuff I had on hand and it's a lot less likely to have the same problem again.