well…
That’s a subjective term, so you will give you my subjective response: I don’t think any of the non-fan driven, non-microprocessor controlled smoke units “burn well”.
By their uncontrolled, unmonitored nature, and the restrictions in design, they cannot in my opinion as an engineer. You are running a heating element open loop with a resistance that varies and unregulated. The only way this is successful is to run conservatively, i.e. so it will not burn out in any condition of “fullness” of fluid. So, since it has to run conservatively to ensure not burning out, volumes of smoke seems out of the question.
Maybe you could experiment with the 18v unit on a 24v system and “dim” the function, but I’m also not real comfortable with PWM output to the heater since it is basically an inductor. (has anyone suggested a snubbing diode yet?) This might let you “tune” the output, and of course turn it off when dry.
That’s my opinion and experience. If you want “good” smoke, I would go with either a self contained system like the massoth, or a “direct drive” system from the decoder.
Best, Greg