I think you are extrapolating one case with a particular radial length to
all vertical antenna ground systems.
The N6LF radial papers detail his NEC-4 simulations and measurements of
vertical antennas and radial systems. If I read his papers correctly base
impedance does track field strength measurements.
Historically there probably have been more cases where base impedance does
not directly allow loss calculations than where it does correlate.
The case where feed resistance and efficiency closely agrees at all numbers
of radials would be very special. One case would be where the radials are
exactly 1/4 wave long electrically, so the antenna base is at a radial
current maximum, and the antenna is 1/4 wave tall. In that case all of the
CONDUCTED ground losses could be directly applicable as a feedpoint
resistance change. Field losses from fields impinging on earth would not
factor in well, but they could be small.
The rule I learned to always apply to theory is where there is an exception,
the theory is no longer correct or valid unless the theory includes **all**
exceptions as part of the theory.
If a ground system has standing waves, which even buried radials do to a
significant extent in sparse radial systems, the impedance at the base is
not representative of the real loss resistance.
I have found that here in several cases, not just one, and others have also
independently reported it without intentionally looking for it. If it is
UNreliable in some case or cases, it cannot be an accurate theory for
general application. To use the theory, you would have to qualify the
exceptions so people could avoid them.
To not use the theory do not require any qualification.
An accurate field strength measurements is always right.
A base resistance reading might be right or wrong..... it might be useful or
Topband reflector - firstname.lastname@example.org