Thank you, Keith, for your contribution. The figure below presents preliminary results obtained after running a script that modifies the aspect ratio of the skeleton slot while maintaining the perimeter of each loop at one wavelength. The upper graph displays the input impedance, while the lower graph illustrates the gain as a function of the aspect ratio (from left to right, the skeleton starts out wide and gets thinner). Note that I use the aspect ratio of each loop, so to obtain Sykes's definition, you should multiply this value by 2.
The reactance curve (Xin in red) remains quite flat, indicating that maintaining the perimeter of each loop at one wavelength (in this case, 15 meters) effectively creates a self-resonant antenna. It appears that for feeding it with a 50 Ohm coaxial cable, an optimal loop aspect ratio is 1.75, which corresponds to a 3.5:1 ratio for the complete skeleton, greater than the 3:1 ratio achieved by Sykes.
The graph of the input resistance (Rin = radiation resistance since I haven't incorporated losses) demonstrates a decrease as the skeleton becomes slimmer. This observation is consistent with the behavior of loop antennas. As the enclosed area of the loop decreases, the radiation resistance also decreases because it is dependent on the loop area. I will soon present this data in a more organized format within an article to better convey the findings.

Warm regards,
Tony