Bob.
Bob,
Best make yourself some coffee; it could be a while.

At the moment I've a lot of things I have to do, also I go back to work for the whole of November (covering for holidays) . I was watching the weather maps starting to turn white over Skandanavia, and we are forecast to have the wind change to the North East this week.
Wind from the North-East chills the air; It will soon be snowing out there.
Also I notice that the US above New York seems to be showing signs.
so I can't do anything with the idea for a while.
But It's there for anyone to play about with. Feel free to fiddle and improve on it.
I think that if you assume a nominal voltage of 13 from the alternator, then (if using separate cells) you would have ten cells. You would only need to apply the grid to one of them.
I don't think that there will be any amplification. If I remember rightly the amplification in a triode comes from the voltage swing across the anode load resistor due to the Anode current variations caused by variation in the grid circuit. A bit like the voltage swing across a transistor collector being down to variation of the base current (except with a triode it's grid voltage). If I also remember rightly there shouldn't be any problem with the grid getting hot as long as you don't let it become positive with respect to what would be the cathode plate.
I think that, once the initial maximum current is established, it shouldn't be too difficult to devise an electronic circuit that tracks the current and increases the negative voltage automatically as the cell starts to heat up and the current tries to increase. So you could maybe do away with the dash mounted variable resistor. On second thoughts, best keep it as a back up.
Anyway, I'm getting too old to worry about making a fortune from this. It's probably been patented years ago anyway (it may not even work).
I'd be interested to read any results that anyone gets trying it.
Dave
(Manta)