Well... I have to admit defeat...
The engine DOES need cooling.
I really don't get it, actually, because I have not seen this engine running significantly over 100 deg C (maybe 110 or so) while I have an identical engine in a plane (also running gasoline BTW) that happily chugs away at 135 deg C all day long, holds full throttle indefinitely and delivers approx 10 times the horsepower this thing does.
So I fitted a 42 mm computer fan (with a pro forma temperature control range from minimum speed at 70 degrees C to full speed at 100 degrees, but I have the strange feeling it will run full speed all the time, so...
).
The fan is currently powered by a separate 3S LiPo, current draw approx 90 mA so a very tiny battery allready suffices, but I intend to change that to a single 3S LiPo supplying fan directly and ignition and RX/servos via a voltage regulator.
Also a tiny modification done to the carburettor (a restrictor in the intake, "calibrated" to do nothing at low loads, but increase intake vacuum at high loads, increasing fuel draw when the engine has to work harder, a necessity for engines running a constant RPM). It looks ugly, nay HIDEOUS, but it works a treat.
It is a simple 1 1/4th" length of 1/8" diameter brass tubing, glued into "something to hold it in front of the (now useless) velocity stack".
And WHAMMO... suddenly the needle could be closed a fair bit more, smoothing out the engine, reducing the torque-roll and full power is now "endless" (it had a very strong tendency to cut out at full pitch, no real problem because it would run OK up to about 90% and I could live with that).
https://www.youtube.com/watch?v=j0Z5EHk_r7IPrior to starting the camera there was a 4 minute warm-up run, where I also leaned the needle a bit.
After the vid I continued for about 4 minutes (all four of them full ahead) and then I checked fuel consumption: 7 ml... on a total tank content of 17 ml, meaning there's about 30 safe minutes in that tank.
I'm happy... I think...