In addition to the comments pointing out other inaccuracies, I think the biggest, most obvious misconception that completely invalidates the whole premise is that axons most certainly do NOT have a single terminal; axons branch and make thousands of individual synapses along their length and at the ends of any branches. For excitatory projection neurons in the cortex, for example, the axons can stretch for many millimeters (or centimeters in larger brains), but also typically branch near their origin to contact the dendrites of neighboring cells.
The number of synapses and number of neurons contacted varies greatly across brain areas, species, cell types, etc.
You might be interested in "small world" organization, such as at link, which I suppose could be considered a form of clustering, but it has absolutely nothing to do with the mathematical distance explanation you are proposing.
The only correct equation you can write here is that the spacing S between neuron A connected to neuron B, must be S<=D+L, where L is the length of A's longest axon, and D is the length of B's longest dendrite; for many pairs, S << D+L. These equations will not lead you directly to any valid conclusions of clustering.