FWIW, on the issue of Informational RFCs seen as cast in stone:
I think I've seen that problem occasionally. I.e. people assigning a far too
high value to a document, just because it is an RFC. The world changes, our
understanding changes, and as Dave pointed out processes evolve… RFCs need to
change now and then as well.
I agree that the issue is much more serious for BCPs. But I think we really
need to get over the issue. The best defense for that problem is to be seen
publishing and revising RFCs. If we start to be afraid of publishing an RFC… I
think we've already lost too much. I do not want to go there.
Just a personal opinion, of course.
Jari