I disagree. Perhaps this is true in already well understood fields of study, but in others we do not have any trunks or branches. Trying to understand general guidelines and principles in such situation necessitates first manufacturing them without much empirical evidence, which seems like a large mistake. In my opinion, the best thinkers are those who try to use general principles and specific details simultaneously, and in turn the best of those ones tend to look at details slightly more often.