I'm not able to comment as to whether or not this draft is ready from a YANG perspective. The amount I don't know about YANG or Automatic Multicast Tunnels is large.  However, from a DNS perspective, this draft is fine.

This document does mention "DNS" a few times, but this only in reference to the mechanism defined in RFC 8777 (DNS Reverse IP AMT Discovery.)  The only relationship to DNS that *this* I-D has is that it has a spot for putting AMTRELAY resource records into the model.  The representation in the model does match the resource record description in RFC 8777.

If there is a concern with the use of DNS in AMT discovery, that issue would be with RFC 8777, not this draft.  Note, that I don't have any concerns with 8777 either.