>I don't see this changing without fundamentally realigning incentives in academia.
I think the corruption in science is a manifestation of Goodharts law. People started to believe everything with the science label blindly. Like if something that starts with "research has shown", then it must be true (now it is "overwhelming evidence"). And whenever that happens, what ever the label is, tends to get bought. (This is why we have toothpaste commercials with people in lab coats).
So the long term solution is to actually decrease the trust of people in "science", the label. Then it will come to a point that just the label of science is not enough to sway the people. Then it has a chance to progress, un-disturbed by these corrupting forces.