Regardless of the method used to prevent a table drop, there should still be regularly practiced drills where you assume a catastrophe of some sort. That means disaster recovery drills where you back up the log, ..., and do a test restore from backup. This just to make sure you have all the steps down like clockwork to reduce data loss and minimise downtime. You also never know when a large back up might be corrupt. You dont want to know this when you actually need it.