To the best of my knowledge, there has _never_ been a
requirement that cited documents be available online, and
especially that authoritative copies be available online.
But perhaps there should be one as to new RFCs.
That means nobody can cite a printed book, or an article from an ACM
or IEEE journal. (To the extent the latter two are online, they're
PDFs behind a paywall.)
This seems to me to be making rules for the sake of making rules, not
because there is a problem that needs to be solved.
R's,
John