JML

This formula

and also Bruce Lightner's formula, assumes a perfect stoichiometric combustion of typical gasoline.
It will give reasonable values most of the time.
The results may differ from the correct ones due to:
1. Some (defective) engines may not have a perfectly Stoichiometric tuning.
2. Some gasoline may have different octane grade and modern cars will automatically tune the mixture ratio.
3. In the warm-up and during throttle change, the combustion is not stoichiometric.

Problems 1. and 3. can be corrected if we have access to the Lambda factor, measured from the exhaust gas:

Fuel Mass = Engine Mass Airflow / (Stoichiometric Ratio*Lambda)

I still don’t know how to correct problem 2. Any ideas?

A note on Diesel engines:
A Diesel engine does not use stoichiometric combustion.
So, a reasonable approach of the fuel mass can only be achieved with the Lambda correction and the value of 14.5 for the Stoichiometric Ratio.
The density of Diesel fuel is 7.03 lbs/gal.

JML