My understanding is that ZFS has no way to expand a drive pool/array
You create another vdev within the pool, I think, I have a Raid-Z1 I can expand that but it's a vdev, TBH I'm getting back into this after using Xigmanas (Nas4Free) for many years with no real issues.
But I agree with what you're saying about expandability with Raid, you mention BTRFS, well there were two raid issues on here recently, one was BTRFS and the other XFS both on top of mdadm. Both used different command line checking and recovery, whilst they worked mdadm does not know about the underlying file system. So doing an fsck /dev/mdX just didn't work.
It reminded me of some research I did in respect of Synology and/or Qnap, they use mdadm to create the array, then create an LVM then format that LMV with BTRFS
But for media storage unionfilesystem and SnapRaid are the suggested way to go as the data is not being changed/accessed regularly + they be expanded and drives replaced easily.
Actually I've been thinking about UKenGB questions and your partitioning idea, so could you use a drive to start as a single drive, add a second, drive format it using the same file system as the first, then from the cli mdadm --create and create a Raid 1 via the cli. Not sure if that would work or if OMV would recognise it, and could you do the same and 'convert' the Raid 1 to Raid 5 via cli after first unmounting the Raid 1.