I don't think you and GP are in disagreement. The scientific method to let greater insights replace lesser ones is precisely to try and fail at falsifying obser

I don't think you and GP are in disagreement. The scientific method to let greater insights replace lesser ones is precisely to try and fail at falsifying observations that violate known physical laws. No physical law is holy for this purpose, although scientists will necessarily have heuristics for what presumed observations they consider good enough to examine closer. Evidently, the EmDrive observations made the threshold! And now we know for sure. Well, close enough, anyway.