Blaming skepticism is an interesting point and makes me wonder. What about the publishing culture? Does it map more closely to incrementalism, or is it just that the surface area of the problem space has grown too vast for the human mind to generally see the bigger picture?

Should empiricism be changed to allow fuzziness and uncertainty within the process? Scientific truth seems like a problem space similar to those amenable to metaheuristics. We don't know the form of the optimal / most true model. I'm not positing this for physics. Mainly systems biology and big data fields.