The problem with this hypothesis is that it treats "linguistic backgrounds" as static, as if language creation was not a basic and routine part of the machinery

The problem with this hypothesis is that it treats "linguistic backgrounds" as static, as if language creation was not a basic and routine part of the machinery we use to picture the universe. But this is not the case. The sciences, for example, would be impossible without the routine creation of new language, and once you admit language creation into the fold as a basic linguistic ability you've lost pretty much all the force of the Sapir-Whorf hypothesis as a limit on our understanding. Linguistic background is a weak and temporary barrier to reaching any particular picture of the universe based on the physical evidence because whenever our linguistic background impedes us we create new language. Shakespeare, famous for his neologisms, is a great example of this process, obviously, and language creation is an area where the sciences and the arts overlap very heavily that so far as I know has never been properly studied.