Brooks somewhere notes that the only/best candidate for a 'silver bullet' is reuse -- and I see that as the backbone of the article, or at least the proposed vision. Essential complexity cannot be eliminated, of course, but it can be avoided -- and that can be done more effectively.