As I see it, Annie, the problem of designing a good galley stove breaks down into four parts:
Choice of fuel. Methane (natural gas) is out because it cannot compressed into a liquid and transported in convenient tanks. All the other hydrocarbons in the "alkane" series (propane, butane onwards evaporate into gases that are heavier than air, something that is tolerable in a house, or when camping, but not aboard. The lighter ones, Coleman fuel/white gas/panel wipe/naphtha, evaporate very fast, and so I reluctantly discard them for use aboard. Similarly petrol/gasoline is not safe enough for use in a cabin. It's only when we reach kerosene in the series that the rate of evaporation becomes acceptably low for safety, and then of course that means that we have to do more work to get it into gaseous form for use. The alcohols, methanol and ethanol, still produce heavier than air vapours, but they evaporate relatively slowly, the % concentration of vapour in air that has to be reached before they will burn is higher than the alkanes and when they do ignite, they burn but don't explode. It seems to me that it's a no-brainer to cook using alcohol.
explosive-concentration-limits
The burner. Making a burner to use alcohol is a matter of getting the fuel/air ratio right. Methanol and ethanol both need less air than the alkanes, but methanol needs air less than ethanol. So a burner has to specifically designed for each -
This will explain why I found the Maxie burner to be unsuitable for use with the methanol that is available in Canada - it was getting too much air. It may also have a bearing on the way that the Maxie burner disintegrates - too much air produces an "oxidising" flame, which is trying to convert the metal of the burner into oxides. Less air, and a "reducing" flame, might give a longer burner life.
The fuel delivery. Gravity is reliable, but variable. A gravity-fed alcohol burner must not have too much or too little head of pressure, and oscillations, or pulsing of the fuel supply, can be a problem. A pressurised tank also is not ideal. This is where a lift pump comes in, as it can be made to deliver a constant flow or constant pressure, or both. I don't think the use of a pump should be discounted. A peristaltic pump, drawing from a low level tank, may not need a valve downstream of it.
Safety. Vapours in the bilge due to spilt fuel are the main hazard to be guarded against. Apart from spillages during refuelling, there are issues with leaks from fuel lines and valves. A gravity tank is not failsafe, and neither is a pressurised tank. Unless a low level tank itself leaks, it is more failsafe than a gravity or pressurised tank: if the fuel line or valvework fails, there is less possibility of major leakage (assuming that if the pump is in operation, then you are actually cooking and would be aware of any malfunction causing fuel to be pumped out through a leak). Again, a lift pump should not be discounted.
For the engineers amongst us, here are two sketches of different approaches that I want to try. The first uses the same geometry as the Maxie burner,the second is a pot burner/perforated chimney burner.