This seems to me like a noisy proxy to simply not doing unimportant things, relying on hindsight instead of judgement to identify them. I guess the technique could work but I prefer actively trying to identify things that might look important but really aren't. In addition I'm not too fond of "communicating" by lack of communication or intentional delays - if I think someone can figure out something on their own I might say "Listen, I'm short on time right now, but I think it's not too hard - try searching for XYZ". Similarly for other issues mentioned in the article such as employee initiative. Again, this technique might work here, but I feel that there must be a better way to encourage initiative than ghosting them.
This is a great perspective. I would at a minimum batch up reading emails so one can get some deep work done. Maybe twice a day is ok for reading emails in general. Depending on the situation it could be more or less. Reading Deep Work by Cal Newport can help figure that out.