Honestly, after seeing how poorly scientists botched modeling covid-19 infection/death rates, I just get how it's possible to model something like this which is several decades in the future and have confidence in your prediction.
But i guess 15 years is not enough for you.
Well, it's simple. Look at BP scientist model in the late 70's and look how much we diverged. 40 years, if the model is false you will see it fast no?
Since you're a skeptic, you have to follow the rule: look at the evidence that say you're wrong before looking at the evidence that say you're right. Luckily, those models can be the evidence of both at the same time, so you only have to do your own research twice!
What's the alternative, to only act when you have 100% certainty about anything? Would doing absolutely nothing between the beginning of 2020 and now have resulted in better or worse outcomes for Covid?