I've done some more work on the aluminum version of the model. This version of the model has gain pushed as high as I can by adjusting the dimensions of the antenna itself. Here it is in comparison to an aluminum dipole reference.
As you can see, it doesn't quite have the 2 dB gain over the dipole reference, but 1.77 dB in gain isn't bad. Another thing to note, I have been keeping track of the efficiency numbers, and this version of the Vector model is actually more efficient than the aluminum dipole reference, although not by much. The AGT is 1.005.
On to currents in the basket area...
Donald, when it comes to the currents and the 1/4 (or in this case near 1/4 wavelength sections) on the radials you may be interested to know that the elements for the inside portion of the radials has a current distribution similar to the 1/4 wavelength section with sources 90 degrees out of phase. You can see the characteristic dip in the colors above for those wires. However, the outside of the radials more closely resemble, although not perfectly, two sources on a wire that are 1/4 wavelength apart, and are in phase with each other. You can't see that here, I had to look at the current distribution numbers of the element directly to see it. The peak is closer to the bottom of the element as opposed to the middle, but there is a rise and fall to the currents. In my opinion, this current appears to match the CST image for the outside part of the radials even more than our test models with sources on either side of an element from before.
Another thing of note with this model, and I noticed this before when I was initially working with the model before the aluminum conversion that Henry suggested. When I lengthened the radial sections I expected to have to shorten the mast length to compensate for the additional length, however, to achieve maximum gain with the radial length modification I also had to lengthen the mast section. This is not intuitive to me as I initially thought to compensate for making an element longer I must make another attached element shorter. There is clearly something for me to learn about how NEC2 is handling currents in this layout.
Just to make a note, yep disclaimer time. This model is an experiment designed to test an idea that has been proposed and see what would happen. I, the creator of the model, am not saying this is definitely the solution to what is happening, but it is a step in a direction. I hope it is the right direction, as it is not my intention to complicate a complicated topic even more. As always, I am interested in any suggestion that might help me create a more accurate model to test this idea, or any known flaws I have overlooked.
And finely, as I mentioned I would above, here is a link to the .nec file from 4NEC2 (I renamed it to add a .txt extension so the webserver would let you have access to it).
The DB
