That's not how science works. If you doubt the result, do your own experiment and publish it. The reputations will take care of themselves.
Obviously signing off on known-fake data is straight up lying, which must remain in a different category than simply doing a study that doesn't replicate.
I think you’ll find that if you actually read the papers themselves, they don’t make nearly as broad of claims as they’re purported to in media or the public consciousness.