Keith said SHOULD means MUST ..., which is clearly not the case,
otherwise the separate terms wouldn't be defined in separate sections
with separate meanings and given separate conditions for use (in
section 6).
No, I didn't say "SHOULD means MUST", I said "MUST do this unless you
have a good reason to make an exception." Which is a paraphrase of the
more carefully worded language in 2119. Of course, the 2119 language is
more accurate than the paraphrase.
In particular, MUST is an imperative and MUST only be used where
there is potential for harm or where required for interoperation.
That language was intended to apply to all of the keywords in the
document. e.g. SHOULD is also an imperative.
It also turns out that there are occasionally other good reasons to use
MUST, SHOULD, etc. which has resulted in a couple of RFCs either
defining these keywords internally or referencing another set of
definitions of these keywords.
It expresses an absolute requirement. SHOULD is a recommendation. A
recommendation which should not be made lightly or ignored without a
valid reason, but certainly not an absolute requirement.
Now I think you are the one interpreting 2119 liberally. It doesn't say
that SHOULD is a recommendation.
Keith