And the definition above doesn't cater to the fact that we change what is known, based on fact. With the definition above, any belief about the this that was unknown is suddenly, magically erased once the fact emerges.

My take is instead that if you believe in something that hasn't yet been proven which turns out to be true, you should still rightly believe in it. If you believe in something that is proven wrong, you should change your belief.

I'd rather say that beliefs are internal and facts external. We don't know (and can't know) all facts. The problem here is you first have to believe that something is a fact, in order to change your other beliefs.

Science is the process of allowing - trusting - others to state facts rather than having to find out all the facts yourself. If we don't trust in science, anything can be true because you can still believe that something is a fact, even if you are wrong.