Walter,

I have not been reading all messages on this thread, but I have some
comments on your item:

If you measured the bandwidth of an entire antenna system, (earth-tuning
coil-antenna), then measured the bandwidth of the coil itself alone, could
this be used to get some idea of earth losses?  Could it be assumed the
ohmic losses in the system are negligible compared with earth?

It is prudent to call them "all other losses" and that includes equivalent
earth loss.  One factor that is easy to overlook is loss within the
transmitter (the source resistance looking back into the transmitter).  This
is fairly hard to measure as it is easy to fry test gear.  I do know from
operation of my own LF PA that the source resistance changes depending on
power level.  Even though the transmitter is optimised for 50 ohm loading
(low pass filter and coupling designed for 50 ohm interconnection), that
does not mean it is 50 ohms looking back into the transmitter.  The more
efficient the transmitter, the lower the effective source resistance (the
idea is to get maximum power in the load).  Transmitter efficiency is a
rough indication of trend, with highest efficiency having lowest resistance.
Hard switching finals likely have much lower source resistance than finals
that are linear amplifiers.

If the same antenna is used for receiving, then it could be that the
receiver presents a good 50 ohm load, in which case the antenna system Q
would be lower than when transmitting (where the transmitter is a low source
resistance).

There are parallel cases with high Q HF transmitting antennas, where antenna
tuners have separate optimum adjustments for tx and rx.  I believe the
explanation is in the differing resistances (or impedances with differing R
component) of the transmitter output and receiver input.

73, Bob ZL2CA