He's probably right. HN has some serious talent around, but it also has as rather core principles a lot of stuff that's deleterious to the kind of understanding that forms the basis of what he considers a good programmer to be.
I have throughout my career tried to maintain a grasp of everything in my stack (I went through a "study the code for HotSpot in-depth" phase that, while I don't regret it, I'm glad is over--though I may be about to start in on MRI...) and I try to make intelligent decisions about when to approach problems at different levels. I see a lot more trend-following and operating off of tribal conventional wisdom than I would like.
