I totally get the argument. But in my experience it's rarely the first idea that you end up running with, or that will work on the market. It's generally sign

I totally get the argument. But in my experience it's rarely the first idea that you end up running with, or that will work on the market. It's generally significantly different project in the end if you're able to get to where the market is telling you to go. This leads to the idea that unless the code is conceived well up front it's highly unlikely you're going to be able to learn from the market and adapt to what it's telling you if you're working with a bunch of ill conceived code from the get go.