You bring up a good point. I've seen issues happen and remain undiscovered for days. The problem is no one realizes you are running in a degraded state and the next failure really takes a toll on the network (proper alerting and monitoring can fix this!).
I guess I like being hands on. I don't mind getting a page at 2 in the morning and having to get up to reroute traffic. I'd rather make sure it was done right and documented the first time. On the other hand, some of this is appealing for certain circumstances.