Editing Openai/6897769e-4ee4-800f-aba5-69cca34f701c (section)

==== - Avoid Planck circularity. If you assume δ=ℓP\delta=\ell_Pδ=ℓP and ν=1/tP\nu=1/t_Pν=1/tP you bake GGG into the assumptions (because ℓP and tP include GGG). That gives consistency checks but not a derivation. ====
* Which energy counts? In GR everything that carries energy/momentum gravitates. If QAT energy is radiative and escapes quickly to infinity, its local gravitational effect is small; if it is stored in rest-like form, it contributes strongly. You must decide whether QAT events produce persistent energy (particle rest energy) or transient radiation.
* Spin & fermions. The scalar-shell model is a useful toy. Real electrons are spin-½; a full theory would model shells with spinor fields (Dirac fields). That can change coupling structure and numerical factors.
* Gauge invariance and stress–energy averaging. To feed QAT results to cosmology you must compute the appropriate average of TμνT_{\mu\nu}Tμν (the stress-energy tensor) for a statistically homogeneous, isotropic universe. Surface-localized shell currents produce multipole structure which averages out at large scales, leaving isotropic energy density + pressure.