Hmm... My main concern was the accuracy of such a map and the conclusions drawn from it.
But if you have accurate fuel consumption data then that would certainly help with the FC (fuel consumption) part of the BSFC graph.
RPM is an easy part, some injector signals carry rpm info too, but plenty of other sources too.
So how accurate is inferred LOD compared to a dynamometer? Here is a little info
How to calculate Engine Load? [Archive] - Innovate Motorsports Forum , if no MAF, then need things like MAP, IAT, and a Volumetric Efficiency table (very much engine dependant), and I'm surprised injection timing isn't in there (I suppose that is more of an output from an ECU perspective).
So accurate lod from MAP seems to take a fair amount of prior knowledge about the engine and some more sensor taps. But I'll buy %80 accuracy overall Julian, that sounds attainable, not sure how useful it is though.
Another consideration if you focus only on the engine parameters you effectively shut out many of the real world variables, which is of course the point of a BSFC map but has a downside i.e. If you are driving around with 7psi in your tires, a map/tps system might be perfectly happy since now the throttle is open wider for a given RPM and VE has increased, but an injector signal would be telling an entirely different story.
I am also biased by the mental trappings of my driving style, either near bsfc or nothing usually, so personally having a hard time seeing the rest of the map get filled in, and I guess you just may be an ecomodder if you cringe at the thought of spending the fuel to fill it in
It is certainly thought provoking though, 'preciate that.