From debian bug #398270

It would be very helpful if one could do
$ iconv -f emacs-mule -t utf8 file
as one way or another, an emacs buffer sometimes gets saved in mule
encoding -- one cannot guarantee it will never happen, and the person
later who discovers the file has been saved that way might not have,
be able to use, or understand the documents of emacs.

One would like to not need a Ph. D. to turn
=98=A8=BB=98=BC=E2
back into
=E5=8F=B0=E6=8E=83.

Actually one way of rescuing the file would be to put -*-Coding:
emacs-mule-*- at the top, possibly with ed(1), before opening it with
emacs. Actually emacs should auto-detect the coding system, but
sometimes there is a byte of garbage so emacs guesses wrong.

Then one can copy and paste the region into a already open other utf-8
file, unless one understands the mule menus.

Anyway, it would still be good to have it in iconv.