New to this site and really enjoying the wealth of information here. Just read Will Rietveld's stove performance comparison and a couple of points have struck me.
Firstly no comment on apparent higher gas use by remote rather than integrated gas stoves. My own experiments , I haste to say crude in comparison, have lead me to believe that very small burners that direct their flame mostly upward are more efficient at heating our small pots rahter than do-nut ring type burners, this may explain the differences here.
In a constant search for lighter weight I have been using the narrow tall gas canisters that are designed to fit into Asian table top gas stoves. These are lighter by 1 to 2 oz c.f. conventional canisters and contain the same amount of gas, and can now be bought c blended fuel. They are also cheaper. Kovea make an adaptor to use thes canisters if you have a stove designed for remote Lindel valve canisters, but it is heavy.
I have been experimenting c stove building and have made a fitting to tap directly into these canisters. The device weighs 1 oz and includes a metering valve which brings the whole weight to 4 oz for remote canister stove , 4.5 oz if pre heating tube fitted.
However I think the most stunning gain is the filling of the "non-refillable" canisters. This is easily done and has enormous advantages.
Firstly there are never any more partly used canisters . There is an endless supply of used once nearly new empty canisters from other backpackers. The price per canister drops to 40 or 50 cents per fill and means gas becomes a very economical fuel. For shorter trips only the fuel needed needs to be carried i.e intentionally part filling the canister. If your BBQ bottle contains blended [ propane + butane ] fuel then cold weather performance is on a par c commercial cartridges.