This document needs to be standards track. There is no 'competition' between
auto-conf vs. dhcp, they serve different communities and need to stand
alone. The historic error in the IETF process was to require both stateful
and stateless to have any network operate. Each approach really need to
stand alone and be fully functional without dependency on the other. The
missing link on the auto-conf side is 5006, and it needs to be standards
track asap.
if that's your argument, why shouldn't every DHCP option be possible to encode in an ND message?
is DNS _that_ special?