Butterfly or rotary barrel carb, my gutfeeling leans toiward the issue not being in the differences between the two (although probably the butterfly generates a bit more turbulence in the intake downstream of the valve), because the Walbroes with a rotary barrel seem to work OK also. I personally do think that the largest contributing factor is the design of the intake system as a whole: Perpendicular and sharp changes of direction IMHO are the biggest contributors: Every separating system, from cleaning fuel oil by centrifuges to sluicing gold in Alaska, it all comes down to differences in density and changes of direction.

But yes, despite the "simplicity and elegance" of methanol/glow, gas and spark are the better option when it comes to reliability and smoothness. The downside is its high demands on mixture preparation. But I am convinced, especially with the way Chris is going about it, we will get closer and closer to nice and manageable solutions.

I recently had an idea for a "not-TX based" solution, but that hinges on whether I can think of a way to incorporate a throttle position sensor in that beautiful carb with integrated solenoid he designed.
The idea is to, instead of a connection to the RX, hook the driver up to tthe TPS, and the driver then has the curve stored inside. I have to still think about the user interface and the method of setting the curve, but once I have a workable idea, I can see if Hans or Dave can translate that into a working solution. There are two obstacles: One is that darned user interface that has to be workable for everybody, the other is that the TX-based set-up allows for a difference in timing between fuel and throttle, and a TPS sensor would make that impossible. And I recently ran into a situation where the possibility to have fuel preceed throttle turned out to be extremely relevant. So I am not yet there.