Great post and interesting discussion. It seems like it ought to be intuitively obvious that "best fit" is more important than "best practice", a point made well by Woolcock and Chapman's post. Sustainable development is all about context. And if context is paramount, then Seth Kaplan's point about placing priority on investing in the local ought to be just as obvious. In other words, to support Kaplan's point: surely it is better to invest in enabling those who know the context most from their own direct experience to acquire the more technical - and easier to learn - expertise on practices they can adapt. But usually the emphasis of development projects is on doing it the hard way: those with the most knowledge of outside practices try to learn the complex and illusive context in which those practices will be applied. Doesn't really make sense. Undoubtedly the reasons for that have much to do with what Steve Golub is alluding to. Investing in the local might make the most sense in the long run. But, ironically, there are a lot of disincentives in the development world to taking the long view -- budget cycles don't cooperate, for one thing.