[mint2]

Take low hanging fruit, get substantial benefit at low effort. Go to perfection, take 25x+ the effort and by the time your done it’s outdated

[domnomnom]

Part of the "benefit" is the software breaking again in the future. Like I said, there is no cost to software failing and even a benefit. If people died when software failed, that 25x effort would probably become 5x dude to efficiencies of people actually working on the problem