>> as a canister stove runs, the canister cools and output goes down. You have to run a stove a long time to melt snow.
Mystery solved! Thanks, Ryan.
If I may be so bold as to paraphrase you in geek speak...
With a longer boil time, the pot loses more heat energy to the environment, lowering fuel efficiency. Theoretically, an ideal stove would transfer all of its heat to the pot (without wasting flame heating up the air around the pot) in the shortest amount of time possible.
I don't use white gas, so I can't do a comparison test. But, in my experience, the Coleman Xtreme Powermax stove does not suffer from the canister cooling problem. So, based on the BTU/oz of the fuels, and general stove design, I'd expect fuel efficiency comparable to or greater than white gas with this stove.
fwiw, I budget 40g (1.4 oz) of fuel per liter to boil water from snow with my Xtreme. But, thats below 7000ft and usually above 15 degrees or so in North Idaho.
I wonder if we could get comparable efficiency out of a stove-on-top canister stove with a suitable windscreen, or copper wire heat exchanger....Hmmm, back to the lab...
PS - much of this winter stove stuff has been previously discussed on this thread:
also take a look at R. Caffin's article: