It takes a while to change conclusions. The important thing is that they do change in the face of evidence. Personally, I do take some risks by changing early with what I consider sufficient evidence. Like Feynman said about the viscosity issue, many will steer towards what is currently the norm but slowly the data will correct.