I forgot to mention, ...
The square root of the dielectric constant in a capacitor is a refractive index of that material. This suggests that capacitance is prismatic and that waves which interact with a capacitor are both refracted and reflected and, thus, spawn two daughter waves for every time that a parent wave interacts with a capacitor. This is how complicated the behavior of capacitors are.

I would guess that the refractive wave is imaginary while the reflected wave is real, and that an overunity circuit would emphasize refractive capacitive daughter waves while conventionally thermodynamic, lossy circuits may be emphasizing reflective daughter waves since Micro-Cap help-file says, ...

"The ideal capacitor acts as a shock absorber and, thus, must not be too low in capacitance lest it fail to perform this task." -- Ergo, an ideal capacitor must emphasize absorption and discharge of wave energy. Reflection is an indication of a combination of absorption followed by discharge.

Refraction of capacitive reactance is the ideal circumstance for us to encourage since it will magnify reactive power output so long as we minimize real power input. Then, somewhere along the way, the imaginary portion of the complex enumeration of reactive power will get squared. And when it does, it will leave a scanty quantity of real power, a build up of negative watts, and a remainder of complex power which will accumulate over time.

This build up of negative watts will not be figurative as in the case of the passive sign convention's definition of the generation of power. Instead, it will actually reflect that definition in the physical world by making it possible for a passive inductor to become an active inductor and generate current without the necessity of engaging a prime mover's external source of energy in order to move that coil through a magnetic field and, thus, defy Faraday's Law of Induction by superseding it with a higher law whose text has yet to be written into our conventional wisdom.

======================

The capacitance of the Earth is estimated to be in the vicinity of 711 micro Farads on the presumption that it is a solid sphere whose outer concentric shell is infinitely distant. But, what if it is hollow with a presumed thickness of 1 kilometer or less?

6,367.5 km = radius of the Earth
https://en.wikipedia.org/wiki/Earth_radius

1,000 km = assumed thickness of the Earth's shell -- probably less than this

"The refractive index of a nonmagnetic material is the square root of its relative permittivity."
https://glossary.ametsoc.org/wiki/Relative_permittivity

Since the dielectric constant, or: relative permittivity, of a material is the square of its refractive index, and the refractive index of granite bedrock is somewhere between 2 and 3.5, then the average refractive index of granite is 2.75 and its square is 7.5625.
http://eodg.atm.ox.ac.uk/ARIA/data?Rocks_and_Conglomerates/Granite_(Toon_et_al._1977)/granite_Toon_1977.ri
for comparison
http://hyperphysics.phy-astr.gsu.edu/hbase/Tables/diel.html

outer surface diameter versus inner surface diameter of Earth's shell
6,367.5 km - 1,000 km = 5,367.5 km

capacitance of two concentric spheres in which the larger sphere is r1 (the outer surface of Earth)

(4 x Pi x dielectric constant of the material x r1 x r2) / (r1 - r2) = capacitance
https://www.toppr.com/ask/en-us/question/the-capacitance-of-two-concentric-spherical-shells-of-radii-r1/

6,367.5 km x 5,367.5 km    34,177,556.25 km
------------------------ = ---------------- = 34,177.55625 km x 12.566370614359172953850573533118 x 7.5625 ~= 3.25 Mega Farads!... or more...!!!
6,367.5 km - 5,367.5 km         1,000 km

Conclusion
Since my simulation requires the presumption of the maximum capacitance of the Earth is no greater than 99nF, this circuit may not give the expected results unless some modification can be devised which overrides the Earth's capacitance?

Maybe a 10pF capacitor should be built into each of the nine or more grounding rods above where each rod enters into the Earth?

And, ... since this circuit tends to develop high nodal voltages ...

For safety's sake, maybe a dielectric goo should be poured into this circuit's project box to solidify into an immersion of a dielectric medium after, first, lining the interior surface of the project box with conductive plates and each plate is separately (?) connected to its own Earth ground?