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.