Professor Hualin Qian of the Computer Network Information
Center of the Chinese Academy of Sciences
I was a guest at CNNIC/CAS in Beijing twice in 2001, where I met Hualin,
and see him from time to time at ICANN functions.
The address space allocated to China is less than what MIT has, and the
necessity, let alone the utility, of forcing encoding into ASCII (any
algorithm), and forcing encoding from a particular profoundly broken
glyph catalogue (without intermediate mappings), for written Chinese in
a specific byte-indifferent standard network infrastructure application,
is not helpful for Chinese network users.
So yeah, NAT-on-a-China-scale is something reasonable to look into, as
is rational relief from, or useful accomodation to, the other problem.
Existance of well-known problems however, does not prove necessity, or
utility, let alone implementation and adoption of, a particular claim
to a solution.