My experience with Esbit is limited but for what it's worth: when I used it on a Sierra summer trip, I used one Esbit brand tab each meal, basic Esbit folding stove (all I had ever seen at the time, 10+ years ago), homemade foil windscreen, 3/4 liter fairly wide stainless pot. The one tab boiled the 2 cups water and then simmered a few minutes while I stirred, every time, in roughly 50-60 degree air temperatures. No data on the water temperatures but straight out of a Sierra lake I expect 50-60 degrees at most.
White gas numbers: I have tracked my fuel usage on several spring snow camping trips. I keep track of total fuel usage for the trip, and the number of nights I find water and thus do not have to melt.
Averages, in ounces by weight:
if melting, for one person - 3
if not melting, for one person - 1.5
if melting, for two persons - 4.5
If not melting, for two - 2.25
The obvious points - if I have to melt I use double the fuel, and two persons use 1.5 times the fuel of one. However, my ski trip partner eats less than I do so he needs less water to reconstitute, partly explaining why the fuel usage for two is not double that for one. I realize priming fuel use plays a part here, but I believe a very small part, simply because my method is to light the stove once for each meal and once only. Also, when melting, I would melt some snow for cold water as well as for the meal, so one should not assume that melting enough snow for a liter and then bringing that to a boil would use double the fuel of just bringing a liter to a boil.
I have also tested my stoves at home to compare them. To do this I started with 12 cups of 50 degree water and brought it to a measured 210 degrees. Fule usage measured included priming. I have both a whisperlite and a simmerlite, and they came out the same, at 43 grams of fuel to do the job. Coincidentally, this is very close to my number for daily usage for one meal if I am solo and melting snow, so it's a reasonable simulation in that is is a single burn of about the same usage, thus having the same ratio of priming to running. I also have a remote canister stove, but it is a Bulin B5, which is notorious for fuel line problems, and when I ran the same 12 cup test, with canister inverted, I noticed it did not run as well as it had in previous testing with the canister upright. So I don't consider that an accurate test, and until I get another remote canister stove I can't compare accurately. I did run this test with my canister top stove - a Coleman F1 Ultralight (since deceased) and it used 34 grams of fuel to do the job. BPL testing in the past has shown that remote canister stoves seem to be less efficient than top-mounts - though why this is the case is not clear, and it may be differences in burner design. (Roger, can you please compare fuel usage of your remote canister stove to the the top-mount stove whose burner head you are using?) At any rate, even if we assume that a remote stove will be as efficient as the Coleman (noted for its efficiency in BPL testing) then the difference is only about 24%.
I should point out that one's kitchen style makes a difference here. If your backcountry culinary methods involve lighting the stove a number of times during a meal, you'll lose ground quickly via re-priming with a WG stove compared to a canister stove. And if you are not practiced, as has been noted by B.G. and others, you'll use more fuel to prime than those of us who have been using WG stoves for 35 years.
And another couple numbers, from other testing I ran:
Whisperlite, to prime and boil 4 cups water, avg. of 3 runs, 16.7 gms
Simmerlite, same test, avg of 3 runs, 17.7 gms
Note that here the Whisperlite beats the Simmerlite, while in the longer test I did that was not the case. I believe it takes more fuel to prime the simmerlite, but it burns slightly more efficiently and thus catches up in a longer burn.