>"Uh, maybe if I were being paid for this, but I'm not. :)"
Jim, Please keep doing what you're doing – it's a great public service! At least it is for me.
I was putting together a spreadsheet with heats of combustions, gas densities, steel/fuel, and fuel/air ratios thinking that people might find it useful. (I'm that sort of engineer while Jim – you're "playing with fire" as a avocation, right? – your vocation is in a different arena of geekdom, I thought.)
Anyway, short answer, since it came up:
Net BTU/lb for the most common BP fuels:
Canister Gas:
Propane 19,944
n-Butane 19,680
Isobutane 19,629
Proxies for Unleaded Gas:
n-Hexane 19,403
Toluene 17,620
White Gas Proxy:
Naphtalene 16,708
Carbide anyone?:
Acetylene 20,776
Alky stove users:
Methyl alcohol 9,078
Ethyl alcohol 11,929
Source: Perry's Chemical Engineer's Handbook, 6th Edition, Table 9-30, Combustion Constants.
I used Net (rather than Gross) values because no one is condensing water vapor into liquid water.
So while I totally agree with Jim that differences in stove design, pot size, wind screens, heat exchangers, etc, trump fuel content BY A LOT (except for hydrocarbons versus alcohols), I'd also point out that ALL canister gases are within about 1% of each other. I suspect a bigger factor is steel per unit of fuel (obviously bigger canisters are better) and how that varies by manufacturer and by fuel mix (more propane seems like it should have more steel for the same storage temperature rating). But, again, steel/fuel isn't as big as stove design and use variables.