How to handle unreproducible RC bugs when the submitter is MIA?

[Frank Küster <frank@debian.org>]

Dear release team,

#297181 is unreproducible, and the submitter has not answered to our
questions for a while.  I am quite confident that this bug is really
PEBCAK, or more specifically a local misconfiguration, or some old
locally installed Emacs lisp files lying around.

I do not think that this bug justifies auctex's removal from testing; on
the other hand, I don't think it would be proper to lower its severity
before we know more.  But we cannot know whether the submitter will
answer again and allow us to decide, and possibly fix or reassign,
before sarge is frozen.  

How should this bug be handled?

I'm cc-ing -devel, because this question seems to be of general
interest.  Please note that I am not subscribed to -release (but get
mail to -devel or the bug address).

Regards, Frank
-- 
Frank Küster
Inst. f. Biochemie der Univ. Zürich
Debian Developer