should someone be allowed to register a header and state in the
description that this is meant to deprecate an IETF standard header?
Mail-Reply-To seems like the ideal example for that question.
Actually, if you read the Mail-Reply-To spec, you'll see that it's meant
to fix the (problematic) de-facto standard use of Reply-To, which is
very different from the (even more problematic) IETF use of Reply-To.
Granted that Reply-To has significant problems.
But heaven forbid that someone should actually try to understand those
problems, and get consensus on a solution, before deploying something
which might make the problem worse.
Keith