Thanks for tackling this Terry. Did you include a patch, i.e. a diff
file? If not, the “patch” keyword does not apply, IIUC. Plain English
suggestions are helpful but they’re reviewed in a different way than a diff.

“The 3.x docs are by design pretty much free of 2.x references. Which is
to say, they are a fresh start with 3.0 as the base.“
True. That said, I would leave footnote 4, since it provides a useful
information for people caring about performance, and may interest them
in digging into the details of the implementation. It may need an
explicit mention “in CPython”, though.

“So I would also remove footnote 5.”
I don’t know. It’s an historical note about an implementation detail; it
does no harm (apart maybe taking space for no gain) and may be
interesting to some people. Does a core dev have an opinion on that
(Georg?).

“Footnote 4 is currently needed because the text suggests something that
is not true.”
I can’t say if the text is inaccurate or just difficultly readable.

“"Mappings (dictionaries) compare equal if and only if they have the
same key, value) pairs. Order comparision raise TypeError."”
I’ll even say “the same (key, value) pairs, irregardless of their
order”. Is the term “order comparisons” used in the doc? If not, its
meaning is non-obvious, and I’d like to find something clearer.

“"Comparisons other than equality testing raise a TypeError." is not
quite correct because 'comparisons' include 'is' and 'in' which do work
as expected.”
I thought “is” was clearly identity and “in” membership or containement
testing in the doc. Can you support your claim?

Cheers