AFF, ONTAP 9.6P5, CIFS, Aggr Encryption, Thin Provisioning. No SnapShots, Snap Policy disabled
From ONTAP, (like "vol show" or "df") I suspect the answer to this is wait. When a large amount of DELETE is done on a NFS/CIFS share/mount, ONTAP takes some time to update all the free-blocks in the system. It should start freeing pretty quick but in severe cases I have seen upwards of 30-60 minutes to finish.
Are you over-provisioned on the aggregate?
It is possible that the assumed free space was not really there.
Once it was deleted from SQLlog, another volume have "claim" it if it was nearly full.
Might be good to look "storage aggregate show-space" and maybe even "vol show":
set diag ; vol show -sort-by aggregate,volume ; set admin