I think a rigorously executed BE has merits. There needs to be a mindset shift that repeating experiments is not "lame" but rather important though (also true f

I think a rigorously executed BE has merits. There needs to be a mindset shift that repeating experiments is not "lame" but rather important though (also true for psychology) and the experiments need to be set up better by and large. The "pop-BE" books are also nicely written and fun to read (Ariely etc.) which helps the discipline from a Fleckian science theory point of view in my opinion. Reading through some of the famous papers and most of the pop books also gave me some inspiration for A/B-tests to run and made me more aware of interesting ethical issues. The "default" label option mentioned in the post is a good example of that. Before reading up on BE I somewhat naively assumed that websites aren't optimized for things like that :) However for the most part I agree that the most interesting ideas will be absorbed into the mainstream. For some reason the mainstream in economics is rather misunderstood by non economists. I had a pretty twisted view of what it was before I started sampling some up to date papers and getting a better idea (I'm not an economist)