Electron g-2 from the APM phtn at Aether-Unit Resolution

This paper calculates the leading electron magnetic response of the Aether Physics Model (APM) from an occupied-Ligamen structural model and Quantum Measurement Units (QMU). The electron occupies a fixed-electrostatic-polarity, zero-winding sector of the five-dimensional dynamical closure space within the six-dimensional chronogeometry. The port triad and temporal complex convention construct a spin-frame algebra on the occupied half-angle lift. A reciprocal propagation pair gives explicit four-component boundary matrices; the mass shell follows within a stated local, memoryless, minimal transport closure with common speed and native electron loading. A same-geometry family with different speeds shows which remaining dynamical choices geometry does not determine. Minimal coupling gives g₀=2, while QMU charge/action closure and source normalization fix the interaction coordinate αₑ=e²8π eₑₘₐₓ². The aligned one-handoff phtn operator factors into a local characteristic operator and a normalized cell-registration multiplier. In the stated minimal continuum completion, the release and recapture factors generated by the registered action cancel the cell multiplier in a complete internal exchange. The leading magnetic-response integral then gives δₑ⁽¹⁾=|gₑ|⁽¹⁾-22=αₑ/2π=e²16π²eₑₘₐₓ², and the QMU response coordinate is E_μ⁽¹⁾=αₑ(1+αₑ/2π). No anomaly-specific coefficient is fitted to the measured magnetic moment. The numerical test uses atom-recoil determinations of the coupling, independently of the electron anomaly. The rubidium input gives δₑ⁽¹⁾=0.00116140973185, compared with the cited experimental value $0.00115965218059(13)$. Prediction minus measurement is +1.75755126×10⁻⁶, approximately $0.151558%$ of the measured anomaly. The cesium recoil input leaves essentially the same leading residual. The result is a quantitative leading approximation within explicit structural, boundary, and continuum assumptions. The remaining difference is an open physical question. No fitted residual correction or imported higher-order coefficients are included, and no experimental-precision agreement or all-orders equivalence is claimed. The package includes the paper, complete LaTeX source, a reproducible numerical script, and the numerical comparison data. The empirical remainder is approximately -0.326(αₑ/π)². This is a comparison scale, not a derived additional interaction or perturbative order.