Some random thoughts:

I would assume that a sizable fraction of .city and .geosomething 
applications were indeed following a superior goods logic. Maybe even 
some large corporate applications. In itself, this is not bad at all. 
Vanity business-models are a welcome change from the old business-models.

I view the motivation behind most other applications as high risk-reward 
low cost endeavors, despite the real costs highlighted by Milton. 
Meaning they chase the old business models, and not new ones.

Applications by the current domain industry mainly trying to replicate 
the defensive registration business model is a net cost on them: I don't 
foresee much reward. Most of them new TLDs are not gonna have any kind 
of an acceptable payback. No doubt spams n scams and speculation will 
contribute to a certain volume to successful applicants, though the very 
idea of expansion should cool down some of the sld-speculation volume 
the applicants get.

The way I view the expansion is that its many old-business-model 
applications was a necessary evil to flesh out the few applications 
intent on finding new business models. Like you say, Google could very 
much turn this thing on its head. I viewed the 185k as a way to 
constrain old-business-model seeker's purported spread strategy of 
diversifying TLD applications, looking for the ones with the high 
payback compensating for others. (I think some of the old applicants 
from ICANN's troubled history with alternates were able to apply for 
free on all the ones they did apply before, right? If that is so, that 
was an *un*necessary evil, whose challenge should have been met squarely 
before the application process.)

And so for the numbers of application in developing countries, it seems 
only normal that ― not knowing what innovative business model might come 
out of this ― there was no point in applying in the hope of benefiting 
from new-business-model first-movers advantages. That no more than 2 or 
3 applications out of Africa did, seems reasonable and expected. I 
wouldn't have expected many African applications to chase the old 
business-model, and also comparatively fewer should have been chasing 
the vanity model.

If the expansion is about finding *some* new business model (and, 
concomitantly, applications) for domain names, I would still contend 
that the good way to go was to enable such an expansion since the 
alternative is, of course, a planned expansion. Such an alternative 
would be even more messy and would not result in the same level of 
"attack" on the old business models. Because I still consider that an 
expansion changes the dynamics of the old models considerably. Defensive 
registration cannot possibly be made out in every new tld, for instance, 
which should contribute to depleting the defensive value attached to 
semantic strings. In other words, I believe that abundance destroy 
defensive value. At the same time, it destroys some of the stronghold 
put over languages and meanings by TM and IP. Meaning, that competition 
on the old business-model will help to scour the old business model.

Nicolas

On 7/13/2012 12:24 PM, Evan Leibovitch wrote:
> On 13 July 2012 11:22, Milton L Mueller <[log in to unmask] 
> <mailto:[log in to unmask]>> wrote:
>
>     > -----Original Message-----
>     > Of course if you and MM are right, and no one in their right
>     mind would
>     > want one of these things anyway, then it might be a waste of time.
>
>     [Milton L Mueller] That is not my position. At all. I see many
>     reasons for a variety of players to have one, and believe that I
>     was advocating that ICANN open the root to new additions since
>     1996 - before there was an ICANN.
>
>     I have always maintained that IDN TLDs in particular refute any
>     claims that there is no need for new TLDs.
>
>
> I agree. My own position is that while the vast bulk of TLD 
> applications are needless extractions of value from the Internet, a 
> handful are genuinely useful. ESPECIALLY IDNs, there should be at 
> least one (and preferably two or three) in every script.
>
> It will also be useful to have some TLDs that were *truly* based on 
> different business models that did not depend on speculators or 
> defensive registrations, or whose differentiator was more than its 
> string being a new category. I look forward to what Amazon and Google 
> plan to do, since they have far different motives for applying than 
> most of the usual suspects. It will be very interesting to see what 
> happens to the business models of all those speculators if Google 
> maintains the path it has gone in other fields, and starts handing out 
> free second level domains to content providers. Indeed, IMO the only 
> silver lining of the entire gTLD application process is the invitation 
> to the likes of Google to disrupt the domain industry (and possibly 
> destroy much of it).
>
> It will certainly be interesting to see what influence these new 
> players, which dwarf the former "giant" Verisign, exert in the GNSO 
> going forward.
>
>     What I am saying is that new TLDs are a species of what economists
>     call a "superior good"; i.e. goods which make up a larger
>     proportion of consumption as income rises.
>     http://en.wikipedia.org/wiki/Superior_good
>
>
>
> In the extreme sense also known in the vernacular as a luxury item, a 
> term which I would certainly agree applies to most new gTLDs.
>
>      To expect poorer and developing economies to exhibit as much
>     demand per capita for new TLDs as highly developed and richer
>     internet economies is just not realistic. That will, of course,
>     change gradually over time as these economies catch up, perhaps
>     faster than we think (given the way our own economies seem to be
>     stagnant or sinking). I just don't believe we can or should
>     force-feed it in order to make outcomes conform to unrealistic but
>     ideologically attractive expectations.
>
>
> I would go a step further and say that most new gTLDs are not just 
> luxury items but deliberate symbols of vanity and status -- adornments 
> that indicate the buyer's ability to afford something that would to 
> most be totally unnecessary (or obviously overpriced). See 
> "conspicuous consumption".
>
> At a technical level, there's not much you can do with <.foo> that you 
> couldn't do with <.foo.com <http://foo.com>>. In conversations with 
> applicants over the past few years, I have been amazed by the number 
> of community and geo TLDs whose primary rationale seems to be "we 
> deserve it" as opposed to "we need it", furthering the concept of "TLD 
> as collective status symbol" and without heed of the needs of people 
> who actually use the Internet.
>
> - Evan
>