I use an extra layer on the front end by storing my wikis on a network-attached storage unit which uses ZFS. Three redundant copies on different drives, with snapshots taken every 15 minutes and saved for two weeks, daily snapshots saved for six months. (Snapshots are copy-on-write for each file, so they take up almost no space unless you’re changing huge amounts of data on a regular basis.) You can restore a file from a snapshot just by opening a special folder on the drive and finding a version you like, then copying it back.
No good if there’s a firmware bug or the house burns down, but that almost eliminates the possibility of a hard drive failure or user error causing any significant amount of lost data – in TiddlyWiki or anywhere else.
If you only want to give this level of scrutiny to TiddlyWiki though, this is probably overkill – IIRC the TrueNAS unit I’m using starts at $999 USD without disks.