Theorem: XVII-A finite-model release

For the newly declared finite-matrix benchmark:

1. the named bulk effect has an exact representative in the algebra of the independently declared conformal-boundary region $B$;

2. the declared complete binary record channel reproduces its bulk measurement channel exactly, including reference systems;

3. the conditional perturbation allowance is (eq:complete-budget);

4. the diagnostic probability is exactly $1/2+\sqrt{3}/8$;

5. the named record does not reconstruct the noncommuting population effect $J$, with optimal scalar error $1/2$;

6. the declared dimensionless boundary data do not identify the background radius within the family of Proposition prop:radius.

The ending is Constructed: exact finite-model operator reconstruction and complete binary readout for this benchmark.

No identification with an inherited VII-A, V, or Em03 observable, continuum field, boundary CFT, entanglement wedge, or gravitational microscopic model is released. Those connecting problems are not discharged by this theorem.

Proof. Theorem thm:main proves the operator and referenced channel identities. Proposition prop:stability proves the conditional budget. Equation (eq:diagnostic-agreement) evaluates the diagnostic. Proposition prop:record-obstruction proves the stated processing obstruction. Proposition prop:radius proves the metric nonidentifiability statement. Each assertion retains its displayed model and input class.