[throwaway4787]

The Popperian vision of science is good enough for students so that their naive mind can enter the field with idealistic perspectives but it's not how actual professionals practice it in real life. Oftentimes there are multiple theories competing, all with reasonable evidence, suggestions, and subscribers. The reason one theory prevails above another is not that the previous' subscribers were definitely convinced by the others' evidence, but simply that they died of old age. [0] The most egregious example is taxonomy: modern-day biologists have long switched to phylogenetics/cladistics to classify groups of individuals, while old school researchers still cling to traditional classification methods (sometimes using retired terms like "race").

In addition, different fields have different standards and methods for what qualified as evidence: for instance, mathematicians want nothing short of absolute proof - that's a given. But biologists are more from Missouri: they don't care that a theory is 'proven', they want to 'see' it [1]. For this reason they put very little trust in simulations no matter how advanced they are. On the other hand, they don't care that much about the mathematical rigor of the tools they use as long as it works out in the end and can be readily confirmed with something you can 'see'.

[0] https://www.smbc-comics.com/comic/how-math-works

[1] Think of it as a 'constructive' view of evidence: everything must be explicit. For instance, if you want to show that trait X is 'genetic' you need to point out an explicit genetic mechanism and show it in action in model organisms. If you want to show that mechanism Y happens in one's cells you ideally need some microscopy to show Y happening live before your eyes, etc.