I'm blessed to work with another DBA who subscribes to the ideas behind the "lazy DBA" philosophy - don't work harder/more than you have to. He set up some phenomenal monitoring and alerting around all of our scheduled jobs. If a new job is created, it will soon be modified to send alerts on failure. Those alerts pull out the relevant details about any failures that occurred and send those along with the alert. We have alerts for drives out of space that contain a brief listing of critical DB files on those drives. We get detailed failure notices for replication if it has issues. We even have visibility into basic stats/items running on our servers over a web interface for when it's needed.

I agree with Steve's point that this is one of many areas that could be enhanced. Andy also makes a good point that MS could just buy/tweak an existing solution without too much trouble.

Of course, having read this I think I need to bug my co-worker about putting some of this info in a blog post. It would be really helpful to people who manage quite a few systems.