Well, I can't be really sure until you retest with fuse #3 out. If everything is wired exactly as the schematics for a non-indicator light have all shown for the last 75 years, then the only positive feed that goes to the voltage regulator is coming from fuse #4. That feed from fuse #3 was only to show the indicator light and excite the VR. If you are not using the indicator light then the change in the feeds to the VR gets everything going. BUT, now that I hear that there might be wiring under there, that is on the green line, of which I am totally unaware, well that puts a new spin on the game.
Did Doug tell you where he mounted that "shunt," which sounds like it is a 10v Smiths regulator just like what you see in my pic. Yes, that would be a prime candidate for failure.
