The paper under discussion is attached as a PDF at the fifth post of this thread.
The diagram on the first page is a 2D representation of a cross-section of a 3D self-intersecting torus. The intersection is shown large to illustrate how three lengths, analogous to the three electron radii, can be seen as part of the geometry of a self-intersecting torus. The discussion in the paper analyzes this better than I can do by repeating it here.
The second diagram in the paper is also a 2D representation of a cross-section of a 3D self-intersecting torus, with a tiny intersection. This is roughly drawn to scale when the three electron radii are used in the torus geometry.
The diagrams also show that there are actually two radii that define the torus, and combinations of them make up the electron radii. The Compton radius is more typically called the reduced Compton wavelength, or Compton length.
Further on in the paper, calculations of surface area are based on the 3D torus, using base units of meters. Initially, what I call the outer toroidal surface area, is numerically close to twice the Boltzmann constant. This suggested some fine-tuning to see if a better match could be achieved. From here on, the calculations probably make the paper difficult to follow. I arrived at the formulations by trial and error - there is no particular theory involved here. Accurate numerical matches were achieved by using a ratio that led to a temperature equal to the Cosmic Microwave Background temperature. This may seem arbitrary, but associating the CMB temperature with the Boltzmann constant doesn't seem unreasonable.
The Inner surface area, the area of the small intersection at the center of the torus, numerically equals ten times the mass of the electron. I consider this to be what I call a numerical artifact. The number must come from related values, but I don't think it can be physically meaningful. Particle mass shouldn't be directly related to area, since particle mass increases at smaller sizes. The exact value was targeted by the fine-tuning, however.
As part of the fine-tuning process, a small quantum of roughly 2.035E-19m radius is deduced to be traversing an orbit defined by the torus. I consider that this may possibly be the beginning of a physical model of the electron.