Rocket launches have to be close to the ideal candidate for COE-driven improvement. Any actual problems are going to be problems that need solving in the short

Rocket launches have to be close to the ideal candidate for COE-driven improvement. Any actual problems are going to be problems that need solving in the short term, and if they aren't solved, there will probably be short term pain. The costs of each launch are high enough that they have to take low-probability/high-severity risks seriously. Other industries have incentives to defer problems into future years, to externalize the costs, or to just plain hide or ignore them based on things like "accepted industry practice" or "within prescribed safety limits". Somebody needs to balance out the short-term incentives by decision makers to get promoted on the back of a press release about a splashy success, because that success may have a metric ton of failure disguised inside.