Some data is more important than other data. By and large, we haven't changed our policies for user backups, and for keeping multiple copies of data (grandfather/father/son) even when we migrated from tape backups to hard drive-based backups. We still keep a some data offsite. 

The most important thing to do during a server crash is to keep your cool. It's difficult to make good decisions when you're under stress, and doubly so if your job is on the line. What changed with our last server crash and _potential_ data loss was actually a better understanding in management over how important it was to allocate the enought money towards backup systems and software, and this helped to build my case for replacing hard drives in our NAS pre-emptively, rather than waiting for them to go bad. Currently we're scheduled to replace/upgrade drives once a year, which should help to keep our storage systems up and running.