Not sure what exactly it was really for, but my guess is not that.
On the other hand, maybe innovation is spurred by letting google own ".dad".
It definitely increases choice. Before the gTLD expansion, you had your choice of the existing legacy gTLDs, your country's ccTLD, and some ccTLDs from other countries that don't have strict location requirements. That's easily under 100 choices. Contrast with now, post-TLD expansion, in which you have over a thousand choices.
There are more opportunities for technical innovation, though admittedly most of the new TLDs so far have been generic open TLDs. But you can imagine all sorts of new things that are possible with closed or restricted TLDs. Some random ideas include: An entire TLD whose registrations are backed by Namecoin or some other blockchain, a secure TLD with really stringent identity verification requirements to cut down on scamming, and a TLD on which domain names never expire and don't have renewal costs (you'd pay more up front for one, understandably). I'm just throwing random ideas out here; this isn't anything we're necessarily working on. But you can imagine lots more possibilities.
I just don't see it as being anything other than a land grab for most of the TLDs google and other companies registered. If you have a trademark on something, or you have a good technical proposal (e.g. .ncd for namecoin backed domain) that you are willing to support then that's fine. .dad, .shop, .buy, etc...
It's just selling out.
I would love to hear technical reasons for their list, https://googleblog.blogspot.com/2012/05/expanding-internet-d...
Anyway, the project is still cool.
We all know running nameservers doesnt require magical unicorn blood and shouldn't be as obscenely priced as is now.