The issue comes down to the conflation of scientific consensus and personal opinion.

What we call 'science' is a method of constructing a model of the world, then trying to disprove the model, getting more and more confident of the model each time an attempted-disproval fails. Because at any moment another test might come along showing that that model is in some way incorrect, confidence in the model is never 100% - but it can approach it pretty closely.

Somehow, people have come to the conclusion that this is exactly the same as an opinion.