I remember seeing a study that showed that people who take notes on paper do better vs people who take notes on a laptop. People who didn't take notes(or take very few) weren't mentioned.
The problem with the authors principles is that they are based on a study[0] that has failed to replicate and was faulty to begin with[1].
[0] Susanne M. Jaeggi, M. B. (2008). Improving Fluid intelligence With Training on Working Memory. Proceedings of the National Academy of Sciences. doi: 10.1073/pnas.0801268105
Now that I reread the article I see that you are right. The authors whole argument depends on the Jaeggi study which has not been successfully replicated after multiple attempts.
I agree that according to the article you linked it seems that the Dual-N-Back game is no different than any other brain training game which seems to differ from the authors view on it.
However, I don't think that debunks the main points of the authors article (the last 3/4 of the article).
And after adjusting for age, gender, race and alcohol use, drivers who tested positive for marijuana were no more likely to crash than who had not used any drugs or alcohol prior to driving.