Reminds me of when I was a post-doctoral student. Who cares about scientific data and evidence? As long as the hypothesis felt good, it's time to slap together a paper and publish, publish, publish! Especially if you can "explain away" all of the various problems with your model when you talk at some conference.