I have found another place where explicit UCD version is used in the docs:

Doc/reference/lexical_analysis.rst:729:.. [#] http://www.unicode.org/Public/6.1.0/ucd/NameAliases.txt

I am not sure how this case should be handled.  The language reference was deliberately written so that it avoids mentioning specific version of Unicode.  For example, PropList.txt in the "Identifiers and keywords" section is linked to the location of the latest published PropList.txt: <http://unicode.org/Public/UNIDATA/PropList.txt>.  This means that as of today, all versions of documentation (3.0 through 3.4) refere to Unicode 6.2.0 version of this file.  This may be misleading for the users of the older python versions.

In the same section, there is a reference top a "non-normative HTML file listing all valid identifier characters for Unicode 4.1."

I would suggest that instead of linking to an external resource we generate a similar table (possibly in ReST format) in makeunicodedata.py and include it with documentation.