The extended double zepp Fig 8, can be seen as following the similar rules of current flow as a G5RV. Its dipole length is more than a half wave length and the transmission line or stub is clearly part of the antenna. In addition this antenna exhibits extraordinarily high gain in a bi directional radiation pattern. (equal to a power increase of 2) An Attribute that definitely falls with our needs to emulate a mono bander.

Following the current flow we have a symmetrical current flow that emulates two dipoles side by side Fig 9, which produces a narrow lobe with power gain. (Fig 9) The pattern formed by a double zepp also has the beginning of extra side lobes that are possibly detracting the overall theoretical gain of the antenna. (They do not reinforce the main radiation lobes)
So here is an excellent, existing, well recognized antenna that we can try out and compare with a loop antenna with outlets. The object being of getting high gain required of a beam with the required multi band attributes.
The loop now replaces the previous current flow provided by the stub and the intervening dipoles which, in essence, retains the power gain of two. Interestingly the side lobes of the double zepp pattern vanished. with little other apparent difference. So it would appear that the radiation from the center loop portion is subtractive to the formation of secondary lobes, a desirable situation. With the addition of the variable capacitor we have made, or also could make the the G5RV antenna into a tunable high gain antenna, staving off for a while the onset of side lobes.
We are now on new ground that looking back, appears to be OBVIOUS considering that we are only using known phenomena without the introduction of what some would call snake oil, or others would call revolutionary or misguided. But the end object still remains, to find a multi band beam with mono band properties