Since evidence based policy making became a thing. We use it for deciding when to plant crops, deciding what medical treatments should be broadly implemented, deciding what water treatment methods are acceptable...it's fairly popular.
> That concept is completely antithetical to what science is at the core (i.e. making repeatable observations to establish fact).
Either what you have established holds repeatably and you are predicting the future, or it doesn't hold repeatably and it's not, by your definition, science. You may be confusing repeatable experiments with repeatable observations, though.
> I understand that observed trends tend to indicate future direction but this push, rooted in FUD, to make consensus of the scientific community equivalent to absolute fact
At this point the observations are enough where it's an absolute fact. The only question is how much, how fast. That's what the computer modeling is trying to get a handle on.
Plus, your argument is equivalent to playing Russian roulette and claiming that the presence of a bullet somewhere in the gun is not an established fact. That's true. You don't know if it might be loaded. But do you really want to find out?