The upshot of my pragmatic Kantianism is just that we should be humble instead of presuming our concepts and models are ever absolutely adequate to their subject matters. Kant's point was that there are conditions of possible human experience which nonhuman things in themselves ("noumena") needn't care about at all, as it were. So there's always a gap between the map and the terrain.

The question, then, is about the nature of objectivity, or about our best conceptions and propositions which clearly work in some sense. I wouldn't interpret an experiment that clearly works as though it doesn't work. The standards are institutional and instrumental. Given science's social goal of empowering our species at nature's expense, scientific methods of inquiry work better than others.