"Richard Clark" wrote Hi Yuri,

This is a most ambiguous "bothering" in that you haven't put any
quantification to what the "error" leads to. No one can possibly
expect perfection, and ±20% is possibly the best accuracy most hams
can expect in measurement. We have all already identified that the
"error" stemmed from an inappropriate application of lumped inductance
in the place of a helix in modeling.

This begs the question: "What's all the fuss over? What's to be
proven? and How do we know when it has BEEN proven?"

73's
Richard Clark, KB7QHC

I think we are striving to improve our accuracy and reflection of reality in
modeling antennas.
We know that efficiency is proportional to the area under the current curve
along the radiator. The "fatter" the curve, the better. This has been
confirmed by the experimental measurements by varying position of the
loading coil along the radiator and use of top hats.
If the modeling program starts with wrong assumption (as we have seen using
lumped inductance) and one uses multiple elements, like in vertical arrays
or Yagis, then the results get skewed and we get wrong "recipe" for the
antenna design.
The biggest benefit would be in properly optimizing antenna design for the
best rejection, F/B, cleanest pattern, which is more critical than just
optimizing for max gain. Especially loaded arrays for low bands would
benefit most.
One could get good indication by comparing say 3 el loaded Yagi design with
lumped inductance vs. loading stubs or solenoid model.
Unfortunately, or fortunately, I am not retired, nor making living from the
RF stuff and my time is limited to be working full time on this. My interest
is to maximize the station and antenna design for contesting so I can try to
cream some records.
So far, it looks to me that this exercise is worthwhile if we can improve
the accuracy of modeling and our understanding of the phenomena.
Looks like lots of antennas would be damaged by the Midwest tornados, the
ugly WX is heading our way.

73 Yuri, K3BU