maybe the white gas stoves are too complicated for him to operate, no idea, but I approached this as Richard Feynman suggests in 'The pleasure of Finding things out", ie, do not assume you know, allow doubt, reject certainty, do not accept bad test results, do not attempt to prove things to confirm your existing bias. Allow the data to guide your inquiry. Having opinions is fine as long as you don't forget that they are an opinion, not a fact. I believe engineers in general are more prone to opinions than scientists, it's easier to just do stuff in engineering then see if it works, I can see the difference. Feynman also reminds us to never listen to 'experts', that's a direct quote from him, and this is why you don't, experts get lazy and stop testing, because they are experts. I'm not an expert so I tested this stuff, and the results are pretty clear.
I only did the whisperlight for old times sake, but the fact is, if you compare a reversible remote canister setup with a whisperlight, particularly where you will use it a long time per day, the whisperlight is equal at least, maybe superior, but definitely almost the same weight, except you are not bound to silly fuel canisters that you cannot fill out of other fuel containers, meaning you need that heavy metal shell for each and every gram of your gaseous fuel.
I did not expect the white gas to be even remotely in the ballpark of alcohol, and it's not, particularly if you use the pounds/mile metric for weight, ie, what you actually carry over the trip miles, beginning to end.
What surprised me here was just how good whisperlight is compared to gas, plus of course, this is liquid fuel and no worries about the cold etc. When you add remote canister, the small weight advantage of canister stoves vanishes for large boil/melt amounts, that's a fact, and you can discover this fact in about 3 test runs of your stove. And, as you noted, you do get a bit better efficiency with practice on the whisperlight type stove. For a 3 night/4 day trip to the snow melting and boiling 8 cups a day, the 15 night/16 day number I list should be close to the consumption.
day/night 1 14 16
15 – IS 455 143 95
15 – PS 560 169 105
15 – 45C 810 602 570
15 – WW32 848 536 488
15 – 4WW32 758 524 488
As you can see, the 4WW32 yields better start and end weight than the 450 gm canister setup, particularly since you'd be using a 100 gram heavier burner probably for the remote canister to melt snow in the snow. Dinosaur, lol, yeah, right, more like a bird that's very well evolved. Good designs are good designs. The svea 123 is I believe also still a good stove.
That's why it's fun to poke at these things now and then, I'd never done an actual efficiency/consumption test on my whisperlight, nor had I ever used alcohol stove type boil then turn off methods, so this is not bad at all. You do only get the really good white gas efficiency when you cook for more than one, or with snow melting from these initial tests however, I guess for 6 cups it would be a touch better still, probably I'd guess around 26 grams or so. Gas and alcohol should burn fairly consistently, ie, 1 gram fuel heats x gm water, as long as the stoves are made right. Certain alcohol stoves are prone to the 'warm then boil too fast and burn inefficiently then peter out' type performance, but the right design gets rid of that problem. Gas burners just turn the gas on and off, so I assume you can take their efficiency and just multiply or divide by 2 or whatever.
I'm actually impressed by the white gas, I was not expecting that, but it figures, some designs and ideas are just very good, and they don't really stop being good just because people want the next trinket to buy, even when it isn't even actually any better.