I'm thinking out loud here. I just wonder whether "industry standard" statistics for new car emissions are meaningful to consumers.
In Europe, we have standard tests used to determine fuel consumption and emissions. I'm going to assume that the tests are indicative of real world conditions and that manufacturers aren't cheating. Both assumptions are true some of the time -- cars which perform highest in lab tests tend to perform highest in real world tests, and most manufacturers have not been caught cheating.
15 years or so ago, I looked at the mpg (km/litre) and CO2 emission (g/km) for cars listed by BMW. I picked BMW because most of their petrol and diesel engines were from the same generation. Multiplying mpg by CO2 emission for different petrol models with different engines, CO2 per litre of petrol was the same. That was pretty much what I hoped: Carbon input from fuel = Carbon output, consistently. I'd have been confused if there had been an outsider. Similarly for diesel engine BMWs, CO2 per litre of fuel was consistent across models.
What is the point of reporting CO2 emission (g/km) when it is the same pollution measure as mpg or km/litre? Petrol engines and diesel engines are different (calorific value of fuels, thermodynamic efficiency) but it is easy to provide a multiplier for CO2 emissions: 100 g/km of CO2 in a petrol car = 100 * Y for a diesel.
I am assuming conventional passenger cars with a modern ECU and sensors, with an efficient combustion process across different environmental conditions. I think that is a safe assumption. Carbon input = Carbon output, because the exhaust catalyser captures almost all non-CO2 carbon emissions. The UK MOT for petrol engines primarily checks CO2 emissions. It would be more useful if it measured other carbon emissions.
Many environmental campaigners have been surprised that their efforts to reduce CO2 have increased NOx emissions. They don't really know their own history. Problems related to car emissions were identified in the 1950s, and both CO2 and NOx were correctly identified as pollutants. NOx created problems that you could see for yourself -- smog, which was associated with California but cars also created the big smogs in UK cities in the 1950s. NOx pollution is also associated with tall buildings, poor air flow and high street temperature -- you can't just blame engine designers.
As I understand the combustion process in an internal combustion engine, you can have any two of three: thermodynamic efficiency, low CO2 emissions or low NOx emissions. So when car manufacturers were told to cut CO2 emissions and had an economic imperative to reduce mpg, the NOx went up.
And we have the other stuff. Soot, unburnt fuel, particulates. They are a measure of engine inefficiency. They are mostly a diesel thing, old engines, which go away when hybrid electric/diesel power plants replace them.