I'm the one who didn't think it was correct. The way cars are tested is with a diluter taking a sample which is then sampled at the end of a phase and then turned into a mass value based on the dilution and an exhaust measurement.
At the end of the test, each phase mass is corrected for distance and then weighted for a final result.
I think the reason for the bag measurement is because the emissions change a lot with the condition, from a cold start to a 70mph cruise, so a constant ppm does not work to cover it all.
I'll have to check with some of the diesel guys to see if they have a feel for a translation between the two.
And it's been g/mi for 40 years now.