"That's one of the criticisms that the peanut gallery often lobs at it: that it's just based on theories."
Speaking from the peanut gallery, more specifically, it's much more Bayesian than people would have you believe. Popular literature and most science instructors are not a very good source of context.
I'd wager the overwhelming majority of the peanut gallery doesn't even understand Bayes, or how to be Bayesian, to even know what to do with science. They hear "theory" and immediately reject it, unaware of its status as a claim of certitude based on decades of observation that increased its probability of being correct toward 1 more than competing theories.
I've tried having conversations explaining Bayes, general probability, and its relationship to scientific theories. Almost always falls on deaf ears. People are incredibly horrible at math--and I don't even consider myself to be that great at it. Their failures at being good with understanding maths flows into being horrible at understanding science, especially that which involves mathematics as integral part of its probabilistic standing in relation to alternatives.
Speaking from the peanut gallery, more specifically, it's much more Bayesian than people would have you believe. Popular literature and most science instructors are not a very good source of context.
The hype does not equal the delivery.