It is, which is why you want to be conservative. If the duplicated code is obviously supposed to be identical, then predicting the future should be trivial. If it's not obvious, then it's a question of "can I conceive of a reason why I'd want to update one and not the others?". And if the answer is unclear, wait a while and see if anything comes up.

This. I wait until the feature is mature, and has been used by actual people. There's no point trying to clean up code that hasn't finished evolving, or that I don't understand fully.

It also means I have a tidy stack of refactoring to do when I'm bored or need a quick motivational win :)