I was able to reproduce this issue in both 9.13 and 9.15. It appears to be a bug in the /application/applications REST API that occurs specifically when you attempt to create a new export policy at volume creation time.
The POST call structured by System Manager sets the value nas.application-components[0].export-policy.name as "none", which is interpreted as a value so it fails because it can't co-exist with the host entry (192.168.0.22 in your example).
I tested it by using the REST API interface (https://<cluster-ip>/docs/API) to craft a POST call with the export policy name value set as an empty array "[]" instead and the volume is created with an export policy that matches the volume name containing the host.
I'm on my phone at the moment, but I'll come back and share some code snippets later. I realize this doesn't help fix it but at least it explains why it is broken. 😄
