It ships when it ships
Predicting when a task will be completed is hard. Engineers have struggled with this everywhere I’ve worked, including me. Some overestimate the time needed to build a certain feature and others might greatly underestimate their actual work load. But is this even a real problem? No, not really.
Not sure who started these artificial deadlines for things, but they make no sense. Even when dealing with an all-hands-on-deck situation, there isn’t a real deadline in the room. The goal is to ship a fix asap. Similar to a soccer game when your opponent leads 0:1 and there are only 5 minutes left on the clock. The team will try everything to score asap. But outside of such pressured situations the team simply focuses on ball possession and making good plays. There is nobody asking the players at what time they will score the first goal. Not before the game and especially not during it.
I believe managers shouldn’t worry about deadlines so much. They should simply make sure that everyone knows what goals the team is trying to achieve, how to get there and ensure the players play well together. A team that works together extremely well can’t help but be super productive. And that’s the only true measure: Productivity.
To clarify: There are situations where deadlines are non negotiable, but establishing deadlines for the sake of having deadlines isn’t an approach that works well. Try establishing goals, because that’s what these things really are.