There shouldn't be an ethical dilemma. If you find out your previous paper was wrong, you can publish a new paper saying so. That's 2 papers for the price of on

There shouldn't be an ethical dilemma. If you find out your previous paper was wrong, you can publish a new paper saying so. That's 2 papers for the price of one(ish). Everyone wins :P