The article is making the argument that people guess structurally wrong. In the same way that it's perfectly possible under a waterfall design system that you m

The article is making the argument that people guess structurally wrong. In the same way that it's perfectly possible under a waterfall design system that you manage to get all of your requirements mostly right from the start, it's also perfectly possible that a family is able to anticipate their needs for a house well into the future before having built/lived in it, but it's a structurally losing game. An example of an alternative approach championed by a lot of people who want to build houses situated deeply into nature is that they'll live for the first 12 months out of a tent, shifting it to various places on their property. By doing this, they get a feel for how the sun behaves, where the wind moves, where animals like to roam etc. and make a much more informed decision of where to site the house. This is much more akin to the MVP/sprint model of building houses.