Vim commands are like a language - you don't route learn combinations, only the individual "words", and you string them together dynamically, like you would do with English words in a sentence.
How is this scientific? All I see is a lot of speculation with no experimentation/testing. Vague similarities to another study don't cut it - we are not monkeys, and the monkeys weren't playing D2/D3. He's just assumed the premise that D3 is less addictive than D2, and the rest is an argument made to fit this assumption. I'm not saying he's wrong, just saying that this is bad science.