QTT Completed-Event Transport and Quantum Coherence

Real-dial dynamics, conditional interactions, and optical interference \\[G-F^{T}GF=L^{T}ML,\\qquad L=0,\\quad FJ=JF\\quad\\Longrightarrow\\quad U_F^\\dagger U_F=I.\\] When does a completed-record transport become a unitary quantum evolution? This paper constructs the amplitude map from a fixed record rule, a static coherent encoding and an oriented real dial. For a measure-preserving source, the matrix on the right is the exact loss into modes outside the retained sector. Zero leakage and dial compatibility yield complex-linear unitary propagation, with no amplitude matrix supplied as a constructor. Source-work and coherent-action selection. Repeated capacity redistribution with regular response at zero amplitude forces quadratic source work. A separate oriented-action refinement law and calibrated loop response select its first-order kinetic form. In a fixed complete physical work-probe chart, the resulting coherent evolution is complex-linear and unitary. \\[F(x)=mF(x/\\sqrt m),\\qquad \\beta_x(v)=m\\beta_{x/\\sqrt m}(v/\\sqrt m)\\quad\\Longrightarrow\\quad \\mathcal A J\\dot x=Hx\\] The implication uses the stated regularity, local-action, common-dial and action-normalization conditions. These concern work and oriented response before target propagation. An explicit two-port allocation uses Artian volume and surface fractions 1/24 and 1/32, yielding H/E_*=[[7,-3],[-3,7]]/96 with no fitted evolution coefficient. Assigning those shares to an actual interaction remains a physical qualification, not a consequence of the geometric integers alone. A nonlinear alternative with the same UEL normalization and capacity envelope fails the proposed refinement law. Finite-depth and derivative bounds distinguish exact mathematics from finite experimental tolerance. The accompanying qualification program specifies work, action and probe requirements; it is not a sealed apparatus preregistration. The extension supplies an exact conditional selection theorem, while a net reduction of independently justified physical postulates remains to be established. Version 2.7 preserves all seventeen prior scientific parts exactly and adds Part XVIII. The new independent implementation contributes 135 symbolic checks and three provenance checks. Its incoming 147-check report is archived, but its unsupplied original program is not claimed as rerun. The full public reconstruction executes 3,425 assertions plus 25 inherited sampling groups; these are implementation checks, not experiments. Existing optical and atomic results, uncertainties, experimental seals and the main book are unchanged. Version 2.6 extends the finite source-to-unitarity construction to interacting classes and an autonomous 64-record source. The controller selecting the contact order is explicitly retained. Source tables and a static coherent encoding determine the amplitude dynamics without an input unitary matrix. A proper contrast sector gives nontrivial amplitude mixing. \\[PE=EO,\\quad P^TMP=M,\\quad PQ=QP,\\quad QE=EJ\\quad\\Longrightarrow\\quad O^TGO=G,\\quad OJ=JO.\\] A positive stochastic routing table on equally sized finite alphabets is necessarily a permutation when it preserves squared capacity on every profile of a state-separating sector. A stronger real-time group theorem and an explicit moving-encoding counterexample delimit the construction. The original contact, leakage identity and free-C4 classification retain their earlier priority. Within this specified source class, an independently supplied unitary-amplitude rule is redundant. Physical selection of the source law, coherent encoding and action allocation remains the next reduction task; qualified external review has not been obtained. This is an exact conditional theorem extension, not new empirical evidence or a certified net removal of physical postulates. All sixteen prior scientific parts are preserved exactly. The new portable verifier adds 896 independently implemented assertions, including 456 local rule instances, 420 equivariant maps, the autonomous controller, the mixing example and exhaustive small-source routing checks. The v2.6 baseline executes 3287 assertions plus 25 inherited sampling groups. Counts describe software checks, not experiments. The original incoming 168-check implementation was not supplied and is not claimed as rerun. The v2.5 calculation evaluates the matched hydrogen 1S-2S centroid from the unchanged source packet. A twenty-term level budget includes explicit radiative, recoil and nuclear corrections. The external QED matching and nuclear inputs, constructor history and shared-input covariance are explicit. No coefficient is fitted to this hydrogen transition. \\[\\nu_{1S2S}^{\\mathrm{matched}}=2\\,466\\,061\\,412\\,376\\,252.4352\\ldots\\ \\mathrm{Hz},\\qquad u_{\\mathrm{known}}=634\\,286\\,637.8\\ldots\\ \\mathrm{Hz}.\\] The central residual against the 2011 measured centroid is -810782.565 Hz (about -0.329 parts per billion), reduced by a factor of 28355.7 relative to the leading Coulomb calculation. The signed compatibility pulls are -0.001278259 against Parthey et al. 2011 and -0.001278232 against Matveev et al. 2013. These are conditional known-input compatibility results, not discovery significances or two independent confirmations. The measured Fermi-scale anchor dominates the uncertainty; unquantified source/model contributions are not assumed zero. The same calculation yields the Lamb-shift specific difference D21 = 8 L(2S)/h - L(1S)/h = 187225891.249 Hz, with conditional known-input uncertainty 126.919 Hz. Exactly common inverse-cube contact terms cancel without fitting them. This is a calculated output awaiting an independent convention-matched comparison, not the hyperfine D21. All fourteen prior scientific parts are retained exactly. The new independent replay adds 118 matching checks, including two point-Dirac/Uehling quadratures, and eight observational-audit assertions. The v2.5 baseline comprised 2391 assertions plus 25 inherited sampling groups. These are implementation tests, not new experiments. The source master formula, title and subtitle remain unchanged; the book and existing experimental seals are untouched. The v2.4 extension added an explicit source Hamiltonian for atomic transition frequencies. Radial regularity and normalizability derive the integer Coulomb levels. Fixed-nucleus Dirac levels, fine structure, electric-dipole selection, recoil and an exactly assembled helium variational matrix are included. The joint leading result is \\[\\frac{\\omega_{21}}{\\Omega_*}=\\frac38\\chi_\\mu\\alpha_0^2,\\qquad\\frac{\\Gamma_{2p\\to1s}}{\\Omega_*}=\\left(\\frac23\\right)^8\\chi_\\mu\\alpha_0^5.\\] The same inherited source inputs now determine both the color and leading decay rate without supplying an observed wavelength or lifetime to either constructor. The Coulomb/Dirac and E1 formulas are established mathematics; the QTT contribution is the upstream source packet, normalization and matched readout chain. The source electron expression retains its geometric-scalar candidate status, and soft matching remains a printed premise. With one measured Fermi-scale anchor, the calculated leading lifetime is 1.596193893 ns, compatible with the cited 1.60 +/- 0.01 ns measurement. The nonrelativistic 1S-2S gap differs from the dated 2011 precision centroid by -9.32269 ppm; that leading-only comparison is superseded for this centroid by the explicitly matched v2.5 calculation above. No fitted access factor removes that residual. The constant's prior GREEN compatibility remains separate from the new conditional matched-hydrogen comparison. All twelve previous scientific parts are retained byte-for-byte, and existing experimental seals are unchanged. The inherited optical result connects the source construction to optical coupling. The inherited photon-edge stiffness supplies the same coupling in Coulomb binding and canonically normalized emission. Explicit integration of the hydrogen wavefunctions gives the dimensionless leading result \\[\\frac{\\Gamma_{2p\\to1s}}{\\omega_{21}}=\\frac{2^{11}}{3^9\\{4\\pi[8+\\pi\\cos(\\pi/8)+\\lambda_\\gamma]\\}^3}=4.04328772872712\\ldots\\times10^{-8}.\\] The complete fixed five-rail correction \\(\\lambda_\\gamma\\) is printed in the paper. No transition strength or contrast is fitted. The atomic coefficient and electric-dipole formula are known results; the QTT-specific contribution is their connection to the inherited source stiffness. Reduced mass and a matched stationary clock transformation cancel from the ratio. Exact inversion and finite-sensitivity theorems specify how to audit the result without turning an observed lifetime into a constructor. GREEN: CODATA 2022 numerical compatibility. The source inverse coupling is \\(137.0359991659975\\ldots\\), compared with the NIST recommended \\(137.035999177(21)\\), a signed pull of \\(-0.523927\\) quoted standard uncertainties. This is a retrospective recommended-value comparison, conditional on the source soft identification, not a blinded experimental selection of QTT. Older recoil tensions are preserved in the detailed audit. A separate published hydrogen lifetime is compatible with the leading reconstruction at its coarser precision; calculated atomic tables are not counted as independent measurements. All ten preceding scientific parts are retained. A general counting-channel theorem separates ancestry loss from mode leakage. An exhaustive 32-configuration coincidence contact supplies an interaction certificate. The composition theorem distinguishes coherence returning from retained memory from the different process that discards memory at each step. The complete optical development includes symmetry selection of the pi/8 spinor character, quadratic capacity, conditional probability and detector results, eleven memory histories, calibrated signal identifiability, higher-order interference and observational tables. Existing experimental targets and seals are unchanged. The electromagnetic extension derives two transverse posit