Paul Mockapetris has written a paper in which he claims that IPV6 presents a danger to DNS. In my opinion most of this is nonsense.
He describes two problems. The first problem is that a DNS server might return an IPV6 address for a server that does not support IPV6. This is clearly a configuration error on part of the administrator, not a fundamental problem in DNS. IPV6 addresses do not magically appear in DNS, they are purposefully put their by DNS administrators. If such an administrator puts in wrong information, that his mistake. Not a problem with DNS. The same thing is true for IPV4. If the server answers an address where nobody is listening, it won't work. So unless administrators are intentionally going publish wrong information no problems should occur.
Compare this to car-navigation software that directs you to drive onto a non-existant road. This will not happen unless somebody put that non-existant road into the database. That's a human mistake, not a fundamental error.

The second problem is that your browser/OS will try to guess the proper name for non-existing names. So if you look for "hotmail", the system will try hotmail.com, www.hotmail.com, hotmail.net, www.hotmail.net, etc....
This has worked for IPV4 for twenty years. It's not a very good solution, but it's not a real problem either. Things will be exactly the same for IPV6, except that it will make more guesses before giving up.

The main problem described is the increased load on DNS servers. This is probably true for some big DNS servers, but not for the vast majority. Most DNS servers don't do much. Most of the time they are awaiting new requests, and they can instantaneously answer.
Even a tenfold increase in DNS queries will go unnoticed on most DNS servers.
It might be a problem for some big ISP's, however these guys are constantly upgrading their systems anyway. Each domain, (and therefore, each ISP) should have 2 DNS servers anyway, and even the biggest ISP's seldom have more then a handfull. Expanding that service is very straight forward, and completly built into the DNS protocol. Just add a new machine, change a few settings and your done. Nothing fancy, nothing that any admin worth his salt should have a problem with.
Even though it might not seem so at first sight, IPV6 simplifies networking a lot. ISP's can save money on routers to compensate for an increase DNS load. (Ok, I'll admit that's wishfully thinking at this moment).
A mitigating factor could be that more and more people are getting cable/adsl-modems with a built-in DNS cache. That can significantly reduce the load on the providers DNS servers.

All in all this feels like an attempt to draw attention to himself.

WIFI is hot. Everyone wants a wireless network. The goal is to be connected everywhere and anytime. It is possible to do that with UMTS, but you will have to pay for it, and it won't be very fast. Wifi is much nicer, but its range is severly limited; you'll have to depend on the kindness of others to provide you with open WIFI networks. Allthough everyone likes to use an open network, few people are willing to run one. Various companies try to solve this problem by creating free-for-members wifi networks, that you can only join on the condition that you share your network as well. ISP's don't like it, and don't allow it. Allthough I have not yet heard of anybody losing his internet connection because he shared it, sooner or later this will become an issue.

It doesn't have to be a problem. The cable/ADSL providers could turn this into an advantage. They should add a wifi router to every cable/adsl modem they rent out. This router should come with software that allows other customers to join the network. The router makes sure the traffic of the guest is separated from that of the host. It would even be possible to temporary raise the speed on the internet connection, so guest usage won't slow down the hosts connection. As an alternative an ISP could even pay/discount it's customers for sharing their connection.
As all the users are customers of the same provider their should be no problem with sharing the network with strangers. If a guest downloads kiddyporn it will be linked to his account, not to the host.
ISP's won't have any issue with this kind of sharing, as everyone is a customer anyway. It would even be possible to create WIFI only accounts, that solely depend on others to provide them with access points.
It doesn't have to stop there. ISP's could agree to share their networks, just like they have peering contracts for the wired internet. The only thing that is required is a strong authentication system, and there are plenty of those around to chose from.

I registered this domain about a week ago, and added some content over the last few days. As far as I know nobody has linked to me yet, as nobody (except for a few friends) even knows that I registered the domain.
But the bots know about it allready. Google was one of the first, and is know indexing my site every few hours.
I probably triggered it myself by having my blog send a notification, as the first entry in the logs not by myself is from blo.gs/ping.php
Less then 20 seconds later the Googlebot was on my site. This surprises me, as blo.gs is owned by Yahoo.
5 seconds after that Blogslive reported for duty, and after another 20 seconds was joined by Ominexplorer. Technoratie joined in at 2 minutes after the initial ping, followed at the heels by Sphere Scout (never heard of them).
Amazingly enough Yahoo is (so far) the last one to show up. 2 days after the blo.gs entry the Yahoo bot finally showed up.

Some might be intimated by those bots, it does give you a Big Brother feeling.
I however think this is the way toward the future. Computers are getting better and better in automatically finding and linking information. These kind of automated response systems will eventually lead to "artificial intelligence", or, a bit more down to earth, "behaviour that is so complex that it is indistinguishable from having human intelligence".