I assume he means that going for a walk provides alternating periods of unconscious processing and deliberate thinking in a similar cadence to that described in the study. I'm not always actively practicing when I'm out walking, there are times wh... Ooh, a squirrel!
It's the difference between focused and diffused thinking.
My debugging breaks are not to think about it. The unconscious is surprisingly good at ordering things.
The learning ones are to think about it though.
My assumption is that there is a difference between actively pushing yourself to think about it by adding new stuff to your brain as opposed to simply digesting the already added.
But it is pure speculation.
"Not to think about it" on conscious level. The point is to let your subconscious process the information instead.
Also getting time to reflect without bombardment of new information.