I remember Paul Randal mentioning during a presentation recording at least one financial institution going out of business due to their systems being unavailable for several days (regulators place tight demands on availability).

If I remember correctly, that was due to mishandling corruption messages and wasting time panicking before initiating long restore sequences.

Best to contact Paul in case he can share more details (http://www.sqlskills.com/blogs/paul/)