I'd like to propose that bench tests for stove efficiency and boil time use water starting just above freezing. I believe this will produce more repeatable results. It also provides a better real-world test (since I'm more concerned with how long water takes to boil on a cold morning than a hot summer day). 32F water just requires water of any temperature and ice. Stir ice in the water for a couple minutes, strain the ice out, and you have 32F or 0 c water temp. No thermometer needed.
It would be ideal to mimic other field conditions in a stove test but would require fancy lab equipment, wind tunnels, etc.
Thanks. I was playing with my new stove the other day and thought this was the best way to approximate field conditions while still at home. The Jetboil consistently used 6 to 7 grams to boil 3 cups of water. Conditions were sea level, 75F air, no wind.

