One Recoil Does Not Identify Dark Matter: Source Non-Identifiability, Cross-Channel Closure, and the LUX-ZEPLIN High-Energy Event under TCGS-SEQUENTION

The LUX--ZEPLIN (LZ) Collaboration reports one event, LZ230616, in a \SI{2.84}{\tonneyear} extended-window exposure. The record is consistent with a nuclear recoil of $248\pm23_{\rm stat}\pm23_{\rm sys}\,\keV$, with maximum local and look-elsewhere-corrected global significances of $3.4\sigma$ and $2.6\sigma$, respectively. It is a scientifically serious anomaly, but not a particle identification: 1710 events survive compared with $1713\pm39$ fitted counts; the expanded analysis is explicitly non-blind; event-by-event ER--NR timing discrimination is inconclusive; and an electron-capture-tail uncertainty is not propagated. The central energy lies in the high-acceptance region, whereas its combined upper $1\sigma$ band reaches approximately $280.5\,\keV$ and therefore crosses the steep efficiency decline beyond the reported $269.9\,\keV$ half-efficiency point. A framework-independent audit treats the data as an accepted point process, so the total count and the empty regions of feature space constrain hypotheses together with the single high-leverage event. At $248\,\keV$, xenon receives $q\simeq0.246\,\GeV$ and a representative Helm argument $qR_1\simeq7.2$, beyond the first diffraction zero. Ordinary coherent elastic spin-independent scattering is therefore strongly suppressed, leaving a compatibility fibre that is broad but physically structured. At least ten primary interpretations appearing within approximately two days instantiate that fibre through endothermic scattering, several Higgsino and scalar-doublet completions, two independent Peccei--Quinn routes, fermionic absorption, elastic spin-dependent $\mathcal O_4$, and an inelastic dark-photon construction. One event does not determine a unique mass, operator, ultraviolet completion, relic fraction, or even interaction topology. The solar channel supplies a stringent but package-indexed adjudication. Pospelov and Ramani obtain a loop-independent floor near $506\,\keV$ because repeated tree-level inelastic slowing terminates at a uranium stall near $0.20R_\odot$, and a robust $566\,\keV$ result when predicted spin-dependent hydrogen scattering completes thermalization. Six Higgsino-oriented LZ treatments occupy an approximate $300$--$371\,\keV$ envelope; the wider published inelastic-splitting census spans approximately $297$--$371\,\keV$. Both are numerically disjoint from the $506\,\keV$ floor, but only the Higgsino subset inherits the Pospelov--Ramani solar map. Wu, Zhang, and Zhu nevertheless advertise IceCube consistency through destructive spin-independent tree--loop interference. Their version~1 preceded Pospelov--Ramani by approximately twenty-two hours, so it could not have answered the subsequently introduced uranium-stall branch; the present disagreement is sequential and tests complete forward packages rather than a detector event or contour in isolation. Under \TCGS{}, Axiom A3 yields a selector-equivalence criterion that is stated but unexercised by the one-event dataset, while Axiom A2 plus readout non-injectivity supplies a conditional explanation of map multiplicity without evidentially selecting the one-origin ontology. The operational framework contribution here is therefore primarily a type discipline that blocks five category errors, plus one new prospective scientific wager: A4's missing-mass application is bound to a finite preregistered multi-scale suite and a nonempty complexity-bounded family frozen before held-out evaluation. The central conclusion is exact: one recoil was registered; dark matter was not identified.