You could have fooled me NB, but maybe it was someone else that is hung up on the .64 idea...as being something really special.
Also, I guess your words above, concerning the Penetrator vs. Sigma 5/8, sort of proves the point this model tends to show. Again, like I've remarked before, "...I see little to no difference."
The full length of 284.25" is what the extended and re-tuned model ended up being...for the overall length after I made the modification to to the top hat model. This is measured from the physical bottom of the radiator to the tip. The full length radiator is made up of two wires in the model, #2 = 72" and #3 = 212.25", as tuned. The top hat model's overall length is 263.25", with a similar match.
I think you appear incredulous about this length not being a .625 or .64 wavelength for CB, but I've explained already. There are just too many little errors in this model for what you wish to happen. The modification to extend and tune ended up at 21" longer. I started with 20" inches, just like my I-10K manual indicates, but the tune took me to 21" inches. I posted the page regarding the top hat, you must have missed it. The substitution ratio for a top hat appears not to be an exact 1:1 exchange. This ratio turns out to be 5% in this case. Does 5% seem to be some remarkable re-occurring value in antenna work we hear referenced often, like the difference between the physical and electrical lengths of a wire, or the end effect?
This is how my tune turned out, regardless of all the math. Most of those antenna length programs on the Internet don't provide the user the ability to adjust the material resistance, the diameter, or the individual tubing lengths if the radiator is tapered, so you when you get a radiator length you still have to tune.
There probably is math available that will predict pretty close for this issue, and you've probably seen it in books and on the Internet. We see such examples of math, and our eyes glaze over in disbelieve that anybody could understand all that. Modeling is probably as close as you or I are ever likely to get...to using such math.