I have just finished doing some tests with my mini Trangia, I used three aluminium pots from the same manufacturer of different diameters, small pot 106 mm, medium pot 127 mm and large pot 157 mm. All of the tests were conducted under the same conditions, ambient 11C, water start 12C water end 95C. At start and before each run the burner was emptied cooled down and filled with 50.0g of alcohol from a container at ambient temperature. The burner was lit in place under the pot as I had to have my temperature probe in place. As soon as water reached 95C the pot was removed and the cap was put on the burner to extinguish the flame and the burner was re-weighed.
Comment: Because of cool temperatures the tests were conducted in the flame was slow to start but after several minutes a full flame was obtained, this was the same for each test.
Small pot: time to raise water 83C, 8m 40s: fuel used 15.8g
Observation: after full flame was achieved the flame came up side of pot 2-3 cm.
Medium pot: Time to raise water 83C, 7m 50s: fuel used 13.8g
Observation: after full flame was achieved the flame came up side of pot less than 1 cm
Large pot: Time to raise water 83C, 7m 47s: fuel used 14.7g
Observation: flame did not go up side of pot
Discussion: Although the results show the larger diameter pots could be more efficient the results are very close and some caution must be applied to these results as only one test was conducted for each pot size at least three test should have been conducted for each pot to get an average to eliminate any errors, I did not have that time. Unfortunately I was not able to take pictures of the flame today.
Below is a earlier picture of my testing facility, I am using a calibrated thermistor/thermistor bridges with a National Instruments USB 6008 DAQ with Labview DAQ software program written by myself. (Temperature measurement is one of my areas of expertise)