Hmm, it gets stranger now. I disconnected the air input hose at the fan clutch actuator to get that problem out of the way while I look at the other one, and I plugged the dangling hose with a pressure gauge. That pressure gauge tracks the gauges in the cab, the governor cuts out just above 120 PSI like it should, and I don't hear any pop-off valves opening. I noticed that I could also hear the air leak from the engine air intake mushroom with the engine running once I knew it was there.
Next, I disconnected the air inlet elbow from the air compressor head so I could see if that was where the air was coming from. I could hear a pfft and feel some air coming out of the compressor intake on each compression stroke... but then I noticed that I could still hear the air leak from the engine air intake mushroom. So, the air is coming from somewhere other than the intake port of the air compressor. So it's good that my compressor may still be OK, but something is wrong elsewhere.
On shutdown, the intake air leak sound continues for up to 30 seconds or so, and then gradually stops over a few seconds with a humming diaphragm valve sound which seems to come from the vicinity of the air tank under the driver side of the cab.
So, now I need to figure out what else could put air pressure into the engine air intake. I checked the vent line that runs up along the snorkel by plugging the end with my finger, and that's not where the leak is coming from.
While I was waiting for air to build up, I investigated my non-functional tach. First I disconnected the tach cable from the top of the electrical RPM sender on top of the injection pump. Cable turns freely, but the driver in the RPM sender wasn't turning. Next I pulled off the electrical RPM sender, and found that it's bound up and the tach drive shaft in the IP is rusted, broken and not turning. So I'll need a new tach drive, and optionally a new electrical sender (I think it's just used for the diagnostic port, and could probably be left out). Oddly, there's an open 4-pin connector next to the 2-pin one that the electrical RPM sender plugs into, and I haven't figured out where it goes yet.