100% agree. The problem is # of tests and test coverage are measureable. Good design isn't (at least not in the short term). There is something to the argume

100% agree. The problem is # of tests and test coverage are measureable. Good design isn't (at least not in the short term). There is something to the argument that thinking about testability helps with design but in practice this often does not help. Yet another problem is there is never time to throw away and build something new. In the life-cycle of most companies/products the available time decreases as a function of time since now you have customers, you have bugs, you have a fatter organization to feed, etc. The time to get the design right is the beginning and you have to do that fairly quickly...