Nick Wenban-Smith writes:
I had thought that the first piece of work for WT5 is a discussion of the definition of what constitutes a geographic name for the purposes of new gTLDs - hence looking initially at the 2012 Applicant Guidebook collectively to see whether those are fit for purpose. The use of 3166 - the standard for codes for the representation of countries and their subdivisions - is not perfect but has historically always been used as reference going right back to RFC 1591 in 1994 "IANA is not in the business of deciding what is and what is not a country".
Personally while I think use of 3166 produces some odd results I have not (yet) heard of a credible alternative as regards country names and the 2 and 3 letter codes which represent country names.
The ISO 3166 standard is not about country names at all. It does not defines the names of countries. It is all about codes. What RFC 1591 refers to is the use of codes as labels for TLDs. When the standard is using names and takes the UN Terminology database [1] as reference for those names. It is a bad idea to use 3166 as a reference source for geographic names. jaap [1] <https://unterm.un.org/UNTERM/portal/welcome>