A Farewell to ARPs: IPv4 Service on IPv6-Only Networks

It's a little buried in the article, but the draft in question is a mechanism meant to go alongside RFC 8950 and https://datatracker.ietf.org/doc/draft-ietf-intarea-v4-via-v6/

I get that this is targeting pain points in the datacenter, but I wonder what it would look like to deploy this on a homelab network. The use of a sentinel address seems like an elegant way to support a subset of devices that don't yet support this. And getting rid of RFC 1918 addresses on my LAN would be one way to deal with the IPv4-prioritized-over-ULA-addresses problem that's also addressed by https://datatracker.ietf.org/doc/draft-ietf-6man-rfc6724-update/

I had a difficult time following this article. It seems to have a lot of fluff words and kinda LLM hints to the prose, but it seems the premise is having a "well-known" address of 192.0.0.11 for IPv4 routers? Hardware/software stacks aware of this special address know to use neighbor discovery to get the MAC instead of ARP? old stacks can still use ARP to get the MAC?