RAID is a way of configuring hard drives (hdd) so that in the case of a hdd failure data isn't lost and the server can still operate, albeit in a degraded state. Once the failed hdd is replaced, the RAID rebuilds that hdd with the data that was on the failed one.
Some RAID configurations allow for being able to lose more than one hdd and still be functional but for most of the common configs, if more than 1 hdd is lost, the whole array and all data is lost.
It sounds like they lost an hdd, replaced it with what they thought was a good hdd but the rebuild process kept failing. One of those failures cause the raid array data to become corrupted. Thus, after sorting out and replacing any of the bad hdds, they would have had to restore from backups.
Sounds like they did what they were supposed to but kept having bad luck (bad hdd, bad replacement hdd, corrupted array). All in all, it sounds like despite all the obstacles, they did a good and quick job of restoring back to the latest data they had.