It's not the upload that is the issue.. even only uploading a few Gb (updating the snapshot) is taking hours.. it's related to the way btrfs is doing things here.. I am still looking. Kopia is hashing and caching for hours and hours.. It's a powerful server I have here.. 16 core.. and it's coping.. but even so.. this is too long to have enhance backups held up. It's related to the way Enhance is using btrfs I think.. as a snapshot is taken when a backup happens and while kopia is working this is held up.
It's like kopia is trying to make a new snapshot of the snapshot Enhance already created after the backup
Were you backing up a btrfs partition in your tests?