I was referring specifically to the article in the OP which attempted to minimize the impact of mobile NOx emissions since it's very small compared to power plant emissions. That's extremely disingenuous since with NOx the top 3 variables that determine the social cost is location, location, location... not quantity. 1mg of NOx in downtown LA is more harmful than 1kg in Montana.
I generally agree, but it's still more complicated than that depending on the ambient chemistry of the location. For example, this well-respected model, "APEEP", shows that in two counties in the U.S., marginal NOx emissions are actually beneficial!
The combined population of those two counties (apparently Orange County, CA, and Cook County, IL) have a combined population of over 8 million.
Remember, NOx destroys ozone under specific ambient chemistry conditions.