Cool insights. I never got very deep into the approach, as it seemed like committee-designed bloatware. Trying to protect up-front against all sorts of possible changes is a premature optimization which loads lots of technical debt onto any large system.
Comments
In practice, that takes so much rigor that I've never seen any project pull it off and enjoy the theorical benefits.