What's the impact? Well, half a million vehicles with the TDI would emit about 0.7 grams of excess NoX per kilometer, or about a gram per mile. So if we assume TDI-equipped vehicles are fairly high mileage--say 20,000 miles per year--we then would guess that those half million vehicles are releasing an extra 10,000 metric tons of NOx per year. Scale it for worldwide diesels, and you have maybe 200,000 metric tons of excess NOx per year released.
In contrast, power plants release approximately 1.5 million short tons of NOx annually in the United States alone. Well, they're not near city centers, right?
Nope. As anyone familiar with the term "transmission losses" might infer, a lot of them are in the city. I'm going to dare to suggest that the EPA is having some serious trouble with the "Pareto Principle" in their regulation of NOx, and that Volkswagen is the least of our concerns in terms of pollution. Concern #1 is, of course, the guys who dumped a few million gallons of polluted water into the Animas River a few weeks back.
Note 2: although I thought that I'd read the EIA statistics on diesel fuel correctly yesterday, a mere 12 million gallons per day for the nation's 5.6 million semis (not to mention other heavy trucks, buses, and the like) seems "a bit low." This link indicates about a 2:1 ratio between gasoline and diesel fuel, indicating about 70 billion gallons per year (191 million gallons per day), and this link indicates that on highway consumption of diesel was about 36 billion gallons in 2012, or about 95 million gallons per day. This would indicate about a million metric tons of NOx from highway diesel under the "old rules" annually, putting it in the same ballpark as power plant emissions.