Interesting and thorough piece in Nature on why basic scientific method of empirical falsification still matters to the integrity of physics however creative the hypothesising.
Of course (as the sole commenter so far says), there are plenty of other valuable kinds of knowledge and evidence, which also require careful reasoning. Some scientists may wish to claim redundancy of any other kind of philosophical thinking – but it wouldn’t be science. (Hat tip to Sabine Hossenfelder on Facebook.)
[Post Note – and quite a few articles responding to the post in Nature, also being collected by Sabine.]