[dcb]

Believe you me, I have thought about it plenty also. There are some major flys in the ointment and that is things like hills and wind.

It all starts with something like weighing the car, so that when it accelerates you can deduce how much (horse) power it took to effect that change in velocity in that amount of time.

Once you know the weight, you can then determine the CDA with some coastdown testing (which is dependant on careful conditions).

So lets say you have accurately weighed the car, and have (perhaps with the help of an onboard device) done the coastdown testing, and have good figures for CDA and CRR and Weight.

So, how do you then, assuming the world isn't flat OR windless, determine that a given peak in bsfc, hasn't been tainted by things like wind or terrain slope or drafts or reduced weight (low tank) or???

Problem with deducing power in the real world is that it doesn't make for a very controlled measurement. Something like coasting in gear downhill and seeing even a tiny bit of acceleration could throw the whole thing off.