I disabled the cache for all shares and moved the tdarr-temp share away from it (which was Cache only), deleted the tdarr-temp folder (even though it was empty), and ran the mover again and it cleared it this time. Moved the folder back to the cache and re-enabled it and its been fine since.
Not sure what happened but for some reason the BTRFS Data "partition" just stayed as this "Data, single: total=16.00GiB, used=15.81GiB" even with multiple balance attempts.
Sorted now though
Since it's late hours I decided to shut everything down and put the array into maintenance.
The NVMe drives showed they were doing a bunch of read/write for about a minute. It seemed to correct the remaining BTRFS errors with the pool.
Scrub started: Tue Aug 16 01:06:35 2022
Status: finished
Duration: 0:23:17
Total to scrub: 1.53TiB
Rate: 1.12GiB/s
Error summary: no errors found