The author often argues over scientifically proven things by trying to reverse them or showing disbelief in the statistics. This article here is a great example.
How many studies of code have been done that only cover cutting new code? What happens to code across time? What are the trade offs between easy to maintain and easy to update/upgrade.
Though not related to programming, a lot of the book's content is about delivering projects on time and on budget, prototyping to fail early, and putting people in the same room to aid communication.