Art, KB9MZ wrote:
"That commercial stations use tines is evidence to me that they serve a
purpose, i.e. electrostatic shielding, not everything, just
electrostatic stuff."

That is exactly correct. The magnetically coupled coils have a rather
broadband response, but the coupling coefficient is less than 1. If you
took the advice of others in this thread and checked Cebik`s pages, he
advises that incomplete coupling (k1) does not insure inefficiency.
High efficiency is available with loose coupling though throughput may
be constricted.

Reduced magnetic coupling may make additional coupling from capacitance
bypassing the transformer throughput more significant. Impedance of the
gratitous capacitance, bypassing the coupling transformer, at the 2nd
harmonic frequency is only 1/2 the impedance of that at the fundamental
frequency. Likewise, at the 3rd harmonic frequency, impedance of the
capacitive reactance is only 1/3 that at the fundamental frequency. The
stray coupling capacitance amounts to a high-frequency boost circuit
which is thwarted by the Faraday screen.

Faraday screening to eliminate priority harmonic coupling to the antenna
is an important advantage in harmonic suppression.
I think the Faraday screen is an expected feature of coupling to the
antenna tower in medium wave broadcasting.

My copy of Terman isn`t at hand now, but I recall that he treats skin
effect, iron-core transformer Faraday screens, and air-core transformer
Faraday screens. Richard Clark is right. The implementations are
slightly different depending if it`s an iron-core low frequency
transformer or an air-core high-frequency transformer. The aim is the
same, eliminating capacitive coupling between the transformer coils.

At short wave frequencies there are other practical ways to rid antennas
of harmonic content. I recall King, Mimno, and Wing describing tuned
transmission line traps to expel harmonics.

Kraus says antenna useful bandwidth is generally a matter of both
pattern and impedance. In a thin dipole, Kraus says the pattern usually
changes slowly with frequency, so likely it`s impedance variation which
limits the useful bandwidth.

In a fat antenna, or in a conical, horn, lens, rhombic, and some others,
the impedance is so well behaved that the pattern variation may limit
the useful bandwidth.

Kraus` 1950 2nd edition of "Antennas" includes several transmission line
tricks for matching short wave antennas over a wider bandwidth. See item
14-24 (Matching Arrangements) on page 434.

Best regards, Richard Harrison, KB5WZI