The underlying facts, if they are public domain, aren't copyrighted
and may be freely used. But originality and creativity, as the court
defined them, aren't limited to the facts. They also apply to the
arrangement and selection of the facts. If you copy the arrangement
and selection you may well infringe the compilation copyright.

"Arrangment", is a non-issue, because in the case of books like MH,
there is minimal authorship involved in the arrangment (order of
chapters, layout of tables, ho hum), but in any case it is trivial to
come up with an different "original" arrangement.

"Selection" is also a non-issue exactly for this type of book, because
the selection of "an encyclopedia of everything known in this subject"
(which is the "handbook" or "definitive reference" claim to value) is
not original. The database model really doesn't apply here, because the
subject is a motley collection of sundry facts, and there isn't some
database antecedent to what is in MH anyway.

I'm also perplexed as to how the old theories of copyright can
survive, because they are based entirely on copies costing something
to make, and money changing hands for copies being made, neither of
which is the case on a.b.e.

How are copyright theories based on copies costing something to make?

That's a conclusion I've made from my own unpublished analysis, which I
can't really detail now. But try this thought experiment: Imagine
anyone could clap three times and *poof* a perfect copy of a book
appears for free, in the privacy of your home. No printing press, no
bindery, just *poof*, it's there. No effort or expense on your part,
just speak the title and clap. Everyone then becomes either a clapping
copyright
criminal, or somebody who drives to Barnes and Noble and spends $25 for
a best-seller, when he could have stayed home and just clapped. An
entire publishing
industry is then based on people doing something much more difficult and
costly than has to be, assuming anybody honors the malum prohibitum of
clapping.

In short, in the ancient idea of copyright (actually only a few hundred
years old), the notion of a tangible "copy" was implicitly something
that took effort and materials (because until recently, it always
did--paper, ink, film, vinyl, whatever), and if someone was going to
that significant trouble and expense, it was the author's right to say
who.

As copying becomes a vanishingly trivial event (in the physical sense),
then the author's right
to control your vanishingly trivial (physical) actions vanishes. As the
lawyers say, the law does not deal with trifles.

Copyright *never* extended to trivial actions of copying: When you read
a book, you are forming images on your retinas that are copies, yet not
protected; you can make these copies all you like. Libraries can
project copies of a single paper book into limitless eyeballs for free.
DVD movies are copied within the electronic player every time they are
played. Such exceptions prove that the basis of copyright is absolutely
not copying per se; it is mere economic expedience, fitted arbitrarily
to each circumstance, and exclusive of trivial
actions, not some high principle of property rights. This is also
exhibited in the 1700s constitutional language of "limited times" and
the more recent doctrine of fair use.

And copying has become non-copying, that is, an altogether trivial
event. A fraction of a square mm of microscopic, invisible, magnetic
polarization patterns change from one jumble of bits to another.
Nothing else in the universe changed. This is a crime?