Fuel heaters have been known for modest gains in the 15% area for years... not saying I believe 62mpg in the stealth though.
Now then, let's look at what debunkers say about things...
http://en.wikipedia.org/wiki/Afterma...mic_Efficiency
Oh snip, that value g appears to be for air at 20C and gasoline as a liquid at 20C... Guess what, if gasoline is fully vaporised as it comes out of the injector at a temperature of near what the coolant is, that value of g changes. (For H2 it's about 5 or 6 BTW) since gasoline is diluted by the air 14.7:1 it changes only a little bit in this case to about 1.35ish, doesn't sound like a lot, but it changes the theoretical obtainable efficiency to around 56%... now remembering we see MPG after all losses, that's a theoretical 20% boost in MPG if we perfectly vaporise the gas, instead of inducting it as liquid droplets, this is due to the SHC of gas going to around 4ish from 2.2ish when heated and vaporised into a gaseous phase.
However, since under low load, typical MPFI and SEFI motors can achieve fairly good vaporisation of the gas anyway in the intake port before valve opening, most motors won't see gains quite this large. Remember though the maximum efficiency numbers would be under high load, with the injector spraying with the valve open, so that's why that calculation appears to be based on liquid fuel at 20C. So if you drive like a maniac you might get the full 20%. Actually, P+G proponents using brisk pulsing might see a decent gain.
Anyone who is awake might notice that a large change in intake air temperature would also affect g quite nicely.
__________________