The analogy I draw from it is not direct, but merely: scientists unavoidably think of their work in certain ways that are themselves not tested. This thinking influences their work.
(You've made some changes since I responded, so let me respond to your last question: No, they cannot. Just as choosing what articles goes into the newspaper makes the news inherently subjective, scientists themselves choose what problems to work on, and experiments to carry out. You want them to just "do the science," but what science? The mental framework that helps them decide what science to do is what Kuhn calls a paradigm, and my claim is that the interpretations of quantum mechanics are such a paradigm. That, then, means that the interpretations are important, even if we can't yet directly test them.)
