Most of Walt's examples are about warfare – we could have learned from the French, but we went ahead and fought Vietnam anyway. And we could have learned from what we experienced in Vietnam, but we went regime-changing in Iraq and Afghanistan anyway. It's interesting that the architects of the latter policies were also involved in Vietnam and therefore the generational argument – that new generations resist internalizing past generations' mistakes – don't apply. Maybe that makes the "we can do better" argument all the stronger?
But there's another idea that begs to be part of this conversation: development. We've been at "development" for 50-odd years – or longer, depending on how you feel about the historical evolution of the whole thing. Maybe you don't think development is a "bad idea" – Walt chose clear liberal targets for a reason, I'm guessing – but it certainly shares the characteristics of other ideas he raises, namely, something we've done over and over that hasn't worked.
A few years ago, Harper's Magazine published a piece about the the first Millennium Village, in Kenya. The piece is buried behind a pay wall, unfortunately, but among the takeaways was this: Sauri, the site of the village, has been a pilot site of Great New Development Ideas before; in fact, the piece says, more than half of the research MVs at the time (2007) were built in places with histories of development projects. It makes some sense – the article says the idea was to avoid inexperience and cultural barriers that can impede development work early on – but it also makes weird science.