I remember this tool coming out of the Qubes community, where people seem to be justifiably skeptical to adding dependencies. Also if this were to run in say Dom0, a shell script is easier to use and inspect than installing applications in the privileged environment that may be stuck on whatever old version your Dom0 environment may be.
So, please keep it as is, continue to do this one thing well, and leave cloud synchronization to the number of other tools already exist to this end.
I wonder if it is useful to consider the duties of backup (which may include version) and synchronization as separate. At least it helps me not fall for the false security in thinking that one system will just do everything.
My backup server has been running ZFS for over a decade and I love it for that and for archives. The bulk of my systems run ext4, because until 20.04 it wasn't really a thing to have bootable ZFS on Ubuntu (yes you could do it, I didn't because reasons). My backups use rsync to pick up incremental changes, so full filesystem scans are involved and the backup server gets pretty busy because of those.
One interesting thing I did at one point was to clone and mount LVs of running Windows VMs, so I could then rsync backup them on the same backup server. With the exception of a handful of physical boxes, my fleet is entirely Linux VMs running on Ganeti hosts with LVM for storage.
This project has me imagining doing VM-level backups by talking to the ganeti hosts and backing up LVM clones outside the host, rather than rsyncing into the host.