Artian Dual-Clock Sagnac Preregistration: A Sealed, Channel-Qualified Test of Metric Proper Time versus QTT Clock Projection

How precise must a dual-clock Sagnac experiment be before it can distinguish two fixed theories rather than merely separate their central values? The ordinary matter-wave Sagnac law is shared by general relativity and QTT. This preregistration therefore compares two physically distinct channels on one ring: a laboratory-referenced matter-wave channel and an autonomous trapped-atom transport channel. Seven operational gates must establish that the second phase record remained autonomous before any theory verdict is allowed. The primary ratio and both predictions were sealed before data: \[ R= \frac{\partial_\Omega\Phi_{\rm AUT}/K_{\rm AUT}^{\rm GR}} {\partial_\Omega\Phi_{\rm LAB}/K_{\rm LAB}^{\rm GR}}, \] \[ \boxed{ R_{\rm GR+QM}=1, \qquad R_{\rm QTT-A1}=\cos\!\left(\frac{\pi}{8}\right) =0.9238795325112867\ldots } \] The targets differ by 7.6120467489%. Neither the projection angle nor an interpolating parameter is fitted. The joint GLS estimator, Fieller interval, \(\pm0.005\) compatibility bands, five-sigma competing-target exclusion rule, seven channel gates, stopping rule, and outcome meanings remain unchanged. Version 1.1 is a timestamped pre-data power amendment. It corrects the earlier prospective uncertainty envelope without widening either compatibility band. No qualifying laboratory data existed or were inspected when the amendment was issued. A live run may now be registered only after a blinded apparatus certificate satisfies \[ \boxed{ \sigma_R^{\rm plan}\leq1.5\times10^{-3}, \qquad P_{\rm class}(R=1)\geq0.80, \qquad P_{\rm class}\!\left(R=\cos\frac{\pi}{8}\right)\geq0.80. } \] For finite simulation, the 95% Wilson lower confidence bound on each classification probability must exceed 0.80. The calculation uses the full blinded bivariate slope covariance and the unchanged Fieller classifier. A failed power gate postpones the primary run and gives no theory verdict. The analytic planning cross-check is \[ P_{\rm band}(\sigma_R) =2\Phi\!\left(\frac{0.005}{\sigma_R}-1.9599639845\right)-1, \qquad P_{\rm band}(0.0015)=0.8303624465\ldots . \] The reconstruction package preserves the original Version 1.0 lock and its prediction hash, records hashes of the unchanged analysis configuration and analysis script, and adds a deterministic two-target power executor, readiness policy, future apparatus checklist, synthetic software receipt, LaTeX source, render audit, scientific audit, and SHA-256 manifest. Synthetic rows validate software reachability only; they are not experimental evidence. No observational result is claimed. Main source anchor: Quantum Traction Theory: Main Book v10.01, especially PDF pp. 554-573. Live definitions: ABC clock, I-clk, and pi/8 projection.