What would this gain us? Networking equipment has had IPv4 built in, not IPv4 and cookies. If we have to change all that, it's about as easy to change to IPv6 (and there's a not of networking equipment with IPv6 built in now). That's been the big problem with any simple IPv4 extension: it would require the scrapping of most of the IPv4 equipment anyway.
The point of V8 was that all existing V4 clients are still 100% compatible -- all of the magic is in the routers and gateways that know about the V8 extensions.
For everything else, as long as the TTL fields returned from the DNS lookups are honored, everything just works.
It isn't scalable. Somebody has to keep track of what is an IP address and what is a cookie on the given IPv4 universe, and clearly we can't have more than 4 billion IP addresses addressed in one universe. We're not going to be able to have all that many IPv4 universes, unlike, say, IPv6 where we can just give everybody an incredible number of direct addresses without problem. IPv6 can handle far more addresses than any version of IPv8, and therefore is far more scalable.
Only the gateways between the universe would need to have tracked what's what, and only to the extent that they are tracking what they issue out. It's as scalable as NAT.
It breaks the net. Suddenly, when I'm following a web link to that Russian site, I find it'll take hours to get there. As far as I'm concerned, that site has ceased to exist.
How do you get to "it'll take hours"?
So, given that it has few if any advantages over IPv6, and works a lot worse, why are you wondering why it went precisely nowhere?
Advantages: 100% compatible with V4.
100% extensible to multiple planet-sized, or country-sized, networks.
Disadvantages: Lacks the concept of a globally unique address.