Isn't it possible to tell your python interpreter to consider by default all .py files as encoded in utf-8??
While this is not good for portability as you may then forget to place the # -*- coding: utf-8 -*-
kind of headers,
it is today a pain for me to have to always have to remember to write this header when writing short temporary scripts with unicode strings inside.
(for example when working on remote envs, I don't use a full-bundled-and-configured editor which could systematically add the header for me, so I have to remember that detail)