No. In the OP's 120 0 120 transformer, I am adding windings in series that are wound in the same direction end to end, to add the voltage. I add the physical reality and the sum is what I get when I measure line to line at the output. The measurements you provide sum to zero, which is not what is found by observation. Your premise fails the test of usefulness (or accuracy).
If I saw the winding in half and stick my head in the middle of the winding, then rotate my head 180deg to look at one winding then the other, the windings will appear to be wound clockwise on one side and counterclockwise on the other. This is the applicable analogue and equal to what your measuring protocol delivers for results. Then you say "subtract them". I say "you first, be my guest". I will help you, use my saw.
Going to a Y source, you are pointing to a different animal in your menagerie, going from a horse to a zebra, and saying 'explain the stripes'.
The Y source example has two iron cores, two phase displaced primary voltage sources, and resulting secondary phi0 = x constant differences. For the OP's example, all solutions reside in a one dimensional coordinate system, linearly along the same line. In the substitution you propose, the solutions reside in a two dimensional coordiate system. You have changed "a" for "b" and wish to claim "a" resembles or is equivalent in some way to "b".
In fact "a" has material differences from "b".