I've thought about this problem before. Thoughts: - If you measure the fuel leaving the fuel tank to the fuel distributor, you must also measure the fuel returning to the fuel tank from the distributor. With two fuel flow transducers, I suspect the cumulative error rate leaves a lot to be desired. BTW ... my 2001 Chevy trip computer is typically a least a 1/2 mpg off actual which turns out to be about 5% error. - You could use 6 fuel flow transducers, one for each injector. The concern here is the low flow rate to each injector and resultant error. - If electrically actuated injectors were used, you could mathematically determine the flow rate via injector open duration. Since the D's injectors are not electrically operated, this won't work. One could map the fuel flow to the mass air flow plate position and determine flow from its position. Again, cumulative error will be the main problem as well as accurately measuring the plate's position. Just some thoughts, Nick