It is my opinion that we need to dedicate some more thought on this matter BEFORE, rather than AFTER introducing new TLDs.
I'm with you. Of the new gTLDs, the only ones that have succeeded so far are .BIZ and .INFO, because they're clones of .COM. Maybe you could call .NAME a success with about 200,000 names. I think .MOBI is likely to succeed because it's going to be put in front of a whole lot of mobile phone users. Other than that, they're all gathering dust. On the other hand, there's no technical reason we can't add at least hundreds of domains per year to the root, and I agree that once ICANN opened the door to new TLDs, it's too late to shut it. The beauty contest approach has been a clear failure, so either an auction or a lottery (which is the same as an auction except for who keeps the money) is the sensible choice. If there's enough TLDs on offer, the price will be reasonable.
For the record, I don't see anything wrong with .nyc, .paris, or .cleveland. The reason is, that it serves a defined purpose - geographical.
Seems to me that ccTLDs already exist for geography, and already solve the problem of distinguishing paris.fr from paris.tx.us. The idea that you need a TLD to be cool is one of the great con jobs of all time.
I think we're a little off-track with the IDN/alias/registry arguments.
Seems to me that there is a pretty strong argument in favor of having .CHINA in Chinese, although the question of whether it's the same as or different from .CN is a significant one. I don't see any reason to treat the Chinese version of .(business) any differently from any other new TLD. R's, John PS: I said most of this a year and a half ago at http://weblog.johnlevine.com/ICANN/whydom.html and other people said it before that.