why, say, a housing boom — which requires shifting resources into housing — doesn’t produce the same kind of unemployment as a housing bust that shifts resources out of housing.
Well, if there's demand for housing, firms in that field are hiring people with specific skills related to building housing, and all the logistics that surround it (home appliances, landscaping, financing). Not, if we know people don't want housing, that merely tells us what not to do. So, if you know that making houses is not profitable, what should you do? That's highly ambiguous.
I don't see why he thinks doing X, and not doing X, are symmetric. There are always many more different ways to not do X, and most of them do not have value (eg, making buggy whips, building pyramids). It's not like people can short housing with their labor. The idea that the 'sets' A and not-A are 'the same' is an idea that has some logic, but shows an insane lack of practical intuition.
What would Krugman do if he could no longer write on economics and politics? I have no idea, and bet he would take a year to figure it out himself.