If you worry about backblaze, the solution isn't to have yet another backup, it's to test your backup recovery process regularly
The risk from not being able to recover your data because of something wrong with your process (like, you are not backing up all the data you need) dwarfs the risk from backblaze to lose your data
That's another type of risk, but if all your backups are in backblaze, you're exposed to "single provider" risk which can become very real very quickly (and not only by backblaze going bankrupt overnight).
Of course, that risk may be entirely acceptable; it's all about risk mitigation and trade-offs, not perfect security.
>If you worry about backblaze, the solution isn't to have yet another backup, it's to test your backup recovery process regularly
The solution is to do both, or all three (or n), where #3 is another backup process. And to test the backups of all processes by trying to read or restore them (onto empty disks) regularly.
I.e., somewhere on a continuum between zero or minimal and the above, depending on how critical your business is.
Because Murphy's Law.
Edited to add:
And how critical your job is (to you).
Of course there is much more to the whole issue, like funds, PHBs, etc.
The risk from not being able to recover your data because of something wrong with your process (like, you are not backing up all the data you need) dwarfs the risk from backblaze to lose your data