But, wouldn't those lights nearer the transformer's side of the circle still be brighter?
I like Gar's method, which, as a circle, would be just like yours, except that one ring would have a break just clockwise of the transformer, and the other would have a break just counter-clockwise of it.
In other words, the lights are supplied via a typical single parallel pair of wires, except one wire gets fed at the end closest to the supply, and the other, via a third wire, at the far end.
This way, every light receives the same voltage, regardless of overall circuit length.
Now that I have been introduced to this circuitry,I ask myself why this is not used in plain vanilla 120VAC lighting.I guess when VD is not a problem the cost of the 3rd conductor is a total waste but when VD is a problem then the added cost of the 3rd wire would be accepted.I guess at this point a comparison of calculating VD to the load center and upsizing conductors as needed is a wash.It would be interesting to see where the break even point was with all factors stirred in.
That would take a bit of calcs but doable,I would first divide the wattage of the lamp by a desired voltage to determine the current,then with that known value I could solve a second equation and determine the resistance of the lamp,you can now substitute different voltage levels and solve for current at that voltage,do a summation of current in resistors in parallel to get total load at a given voltage,then using that solve for VD allowed for the threshhold voltage, which would be the acceptable voltage required to give you a satisfactory light output.A lot of basic Ohm's Law math.
I think I said that right,if not someone will correct it LOL'