Computers are great writing aids but I am continuously frustrated by them for textual research purposes. They cannot model what occurs when I use pencil, paper, and write in the margins. The experience of Kimbro (the man in the article) reminded me a bit of Ted Nelson with his book Computer Lib. Ted couldn't write the book using a computer because it wasn't possible for computers at that time to layout such a complex book. I think this article raises similar issues.
Second, thinking deeply is a challenging exercise which is often ignored for more pressing concerns. Thinking, writing, teaching, and asking questions are the basis of learning. Reading and listening can only take you so far. I often find I do not truly understand something until I can either write about it or teach it. For algorithms I experience this most acutely: I never understand them until I have implemented them.