That's a slippery slope. Where does one just ignore the manuscripts and substitute whatever, one wants to be there?
Very definitely a big question. And I think the answer has to be something like taking some texts as more certain than others, and then holding those where there is a question to a test of "Is it something that is consistent with the rest of Epicurus said?" and "Did other ancient critics (of whom there were many who had lots to say) point out an alleged inconsistency?" But in the end there is no way to be omnisciently "certain" on these questions than any other, and we all have to do the best we can.