It’s a fundamental tension in the scientific method. We create models for how reality works. We _know_ that they are imperfect and incomplete, but at times we forget and therefore treat an idea or even observation that violates the model as “impossible”.

Multiple observations ultimately dispel this failure but single observations are easy to dismiss. This is also why crazy but ultimately correct ideas are difficult to distinguish from crazy and wrong ideas.

And academics spend a lot of time teaching people to understand the models, and are always looking for errors that arise from an incomplete or incorrect understanding of the models. So, it is very natural to treat new theories that violate an incomplete model the same way. Especially in fields like medicine where it is nearly impossible to truly isolate single factors and outcomes.