I think the concern was if you didn’t know that up front. It is pretty safe to split along concerns like suggested. But arguably there is a potential cost in prematurely setting boundaries as it can lead to more complicated code if you ever have to change those. But it is a sort of damned if you do, damned if you don’t scenario. On one extreme you break down stuff so far that you are on the endless hunt of making more components of components and while things are resusable the effective savings are minimal and the required understanding significant. On the other extreme a lot less is reusuable as there is less flexibility but it is much more disposable and easier to manage. Obviously it isn’t just a simple there is one correct decision. I think it’s recognizing there are tradeoffs in any extreme.