It has a real clunker in the middle of an otherwise competent piece. Here is the clunker.
"2. Vastly increased complexity: in the past, it used to be possible to do a few simple things to fix car engines – such as DIY (do-it-yourself) cleaning the spark plugs to improve running. Now, car engines are so complex a computer engineer is often needed to fix problems when they (far more rarely) arise. The question is can test regulators understand engines to the same level of complexity as the best auto engineers? If not, engine performance regulation is a virtually impossible job and something like this could happen again."
To be a regulator in the first place, you have to be persistent, socially skilled and optimistic enough (alternatively, foolish enough) to believe that it is possible to create rules that capture public values and then to enforce those rules. Not surprisingly, regulators tend to focus on black box/end point testing for enforcement purposes, and they also have a tendency to create the simplest testing regime possible that accomplishes their goals.
The clunker here is that it confuses the vague and largely useless concept of complexity with the much more useful concept of testing as part of an offense/defense dynamic. "Complexity" and phrases like "understand engines to the same level of complexity as the best auto engineers" are designed to make people throw up their hands and go, it'll never happen! Boooo! What VW has actually done is exploit an oversimplification in the testing regimen. It wasn't that hard to detect the exploit: slap a cuff on the tailpipe and then measure what's coming out while driving around vs. put the car in a rig and then measure or (worse!) have the automaker install self-diagnostics and periodically dump the data. _Activist_ groups were able to deploy the reality test and were able to repro the results enough to pressure the regulators to replace their oversimplified testing regime with one that better matched Real World Use of a car and also detect the problem. From there, figuring out how the exploit worked was something you handed off to "the best auto engineers". But spotting the results of the exploit in no way required "the best auto engineers" or really _any_ understanding of how things worked in the car.
Now, if you _did_ know how diesel engineers were fixing the nox problem in other makes, you could have looked at a parts list for a VW diesel and go, wait, they are missing this whole subsystem (mumble urea mumble turn nox into nitrogen and water mumble). How are they fixing the nox problem? And then you maybe would have known that something funny was going on and you could have sicced "the best auto engineers" on it -- or just slapped a cuff on the tailpipe and drove around with a testing rig.
_Testing_ people know this stuff. It's not _easy_, but it's not as hard as this article might lead you to believe.
Besides, if you need crazy smart people to measure stuff, we're all fucked anyway. If it's important, we have to simplify it down to us norms.
Regardless, I'm happy to see that some of the more immoral/amoral market theorists are finally caving on this whole idea that values don't matter to the market, only Profit does. It's no good making a Profit, if the entire population is chasing after you with pitchforks, tar, feathers, etc. Profit is money, and money is the reification of Other People Collectively Owing You. If everyone hates you, the IOU is probably not collectible.