This has all been very fun to read, lively strand to say the least!
But I am curious about one thing, Donald; How does this Vector design defy vertically-stacked collinear gain theory? In this "non-apparent collinear" the alleged current maximas aren't spaced the 1/2 wavelength apart needed for 2dBd?
The graph shows maybe .5dBd for 1/4 wave spacing of 2 full 1/2 wave dipoles, but the lower one of these two in your Dominator & the Vector design is only HALF of a 1/2 wave dipole.
At what distance were these tests performed?
I wasn't aware that this design defied collinear gain theory. For if it did all collinear antennas that used the common folded phasing stub would also be defining collinear gain theory since they do not add any extra spacing between elements either. I also don't believe all of the gain comes from the simple idea of a 1/2 wave over a 1/4 wave. Far too much change in the distant signal can be seen by making small changes in element lengths for me to ignore the likelihood that beam tilt is also possible with this design.
Simply lengthening the radials forming the basket causes the signal feeding the upper 1/2 wave to be delayed with respect to currents on the cone. Bob did this, I did this and so did Sirio a few years later in the new Vector. Just a few degrees of delay and we have downward beam tilt. This is a change that cannot be seen locally. All of my testing was done with a vertically polarized FM station transmitting approximately 30 miles away.