I cover this in a Resource:
As for your HDD heads crashing into the platters, I can’t say, other than it would be highly rare event. Especially for Enterprise grade disks. And if it did occur, probably few or one HDD. Hopefully under your redundancy limit, (like RAID-Z2’s 2 disk lost and no data loss).
Next, you probably don’t want a configuration backup performed now. An older one would probably be both better and fine.
One thing people have done is written a script to perform a cron’ed automated backup of the configuration to the data pool. Like once a day and keeping the last 30 days, (or more). This means if your boot pool is dying and has been for 7 days, you can get a configuration backup from 8 or 10 days ago that is expected to be clean.