# n177 — C3, first half: the universal coupling derived; and the
#         great convergence — one room left
# (2026-07-11. The third construction, half done, and the discovery
#  that the remaining halves of everything are the same calculation.)

## The universality theorem (exact, one line)
The action's vertex rewrites identically:
   sum_a dk_a (grad-eps . d_a)^2 = [sum_a dk_a d_a^i d_a^j](d_i eps)(d_j eps)
                                 = h^ij T_ij .
GR'S COUPLING FORM IS THE SHAPE OF THE k(grad eps)^2 TERM. One
vertex, one normalization, every polarization channel - universality
is not assumed, imposed, or tuned: it is an algebraic identity of
the unified action. (This is what Weinberg's theorem will eventually
be entitled to assume; here it comes out of the Lagrangian first.)

## Lattice fidelity (measured)
The 9-direction complex's 4th-moment tensor is isotropic to 10%
(cubic residual; the 3-axis set alone would give ~33% - the face
diagonals do real work). The h.T form therefore holds with a
10%-level cubic correction - a truncation artifact that shrinks
with richer direction sets, named and quantified.

## The matching condition, and THE GREAT CONVERGENCE
Newton rides the constrained trace channel (stiffness S-class);
radiation rides the marginal TT channel (kinetic A, gradient B).
One dimensionless ratio decides whether both carry the same G:
   rho_G = B_TT / (S_trace x f_tensor) = 1   (GR-equality)
And rho_G is computable ONLY in the pinned one-frame state - the
SAME calculation that already owes: Sigma_shear = 0 (marginality,
n145-M2/n153), B/A = c^2 (the speed, n149), and C1's positivity
re-run. EVERY REMAINING DOOR OF THE RADIATIVE SECTOR IS ONE ROOM:
the pinned-state computation, with four numbers due from it -
   Sigma_shear = 0 · B/A = c^2 · rho_G = 1 · A > 0 (re-run)
- each with its executioner already in print (the double pulsar at
1.3e-4 judges rho_G; GW170817 judges B/A; the m_g bound judges
Sigma; C1's suite judges the norm). The program's entire remaining
radiative risk now has a single name, four outputs, and four judges.

## C-list status
C1 done (bubble) · C2 done (helicity +-2, exact + covariant) ·
C3 half done (universality: THEOREM; normalization rho_G: pending
the one room) · C4, C5: their numbers ARE rho_G and its tensor
factors - they collapse into the same room · C6: after.