PUH Theorem 333 (Four Exposures, Three Closed) — T185's ω_BD = 405 Gives γ = 0.9975 Against Cassini's 1 ± 2.3×10⁻⁵, a 107-Fold Violation Traceable to One Identification; While Frame Dragging, Preferred-Frame Velocity and Global Rotation All Pass

Photonic Universe Hypothesis (PUH) — Observational Audit. WHY THIS AUDIT WAS NEEDED. T185 derives ω_BD = (3/2)(K_total/Q(1−Q)) = (3/2)(60/(2/9)) = 405 from Casimir weights, and states in the same sentence that this is "well above the minimum of 40,000 required by Cassini." FOUR HUNDRED AND FIVE IS NOT ABOVE FORTY THOUSAND. The paper understood its own difficulty: two sentences later it calls 405 "a specific, falsifiable prediction of PUH that requires either confirmation at this value or revision of the K-weight formula," and offers an escape — one does not offer an escape from a bound one believes has been cleared. The arithmetic in that sentence contradicts the physics the paper knew, and nothing in the archive has revisited it in four months. THE MEASUREMENT, VERIFIED AGAINST ITS SOURCE. Radio tracking of the Cassini spacecraft passing behind the Sun in 2002 gave γ = 1 + (2.1 ± 2.3)×10⁻⁵, consistent with general relativity's γ = 1. The same measurement yields ω_BD > 40,000 at 95 percent confidence, obtained by comparing the derived Shapiro-delay expression with the measured value. Both figures were checked against the primary literature rather than recalled. γ IS THE QUANTITY BUILT TO DETECT EXACTLY THIS KIND OF THEORY: in the measurement's own framing, bending and delay are proportional to γ + 1, and γ − 1 measures the degree to which gravity is not purely geometric but affected by other fields. THEOREM 333.1 (the exposure and its size). A scalar making the gravitational coupling vary contributes differently to the time part of the metric than the space part, so γ = (1+ω)/(2+ω) drops below unity. COMPUTED: ω = 405 gives γ = 0.997543 and |γ−1| = 2.457×10⁻³, which is 106.8× Cassini's allowance; ω = 4,000 gives 10.9×; ω = 40,000 gives 1.1×; ω = 400,000 gives 0.1×. THE VIOLATION IS A FACTOR OF 107. In physical terms, light grazing the solar limb deflects 1.7515 arcseconds under general relativity and 1.7494 under T185 — a difference of ~2,000 micro-arcseconds, small absolutely but a hundred times larger than the uncertainty of a measurement made to one part in 43,000. The table also shows why the bound is quoted as ω > 40,000: THAT IS γ'S BOUND IN DISGUISE. THREE FURTHER EXPOSURES, ALL OF WHICH CLOSE — reported at the same length as the one that does not. (a) FRAME DRAGGING: the Sun's rotation adds a gravitomagnetic delay, but at Cassini's impact parameter the ratio to the Shapiro delay is 2.2×10⁻⁸ — A THOUSAND TIMES BELOW THE MEASUREMENT NOISE. It neither rescues nor threatens anything. (b) PREFERRED-FRAME VELOCITY EFFECTS: T185's title claims a universal rest frame, which would ordinarily expose α₁ and α₂ (bounded at 3.4×10⁻⁵ and 1.6×10⁻⁹). BUT ITS THEOREM 1 STATES THE CRITERION IN TERMS OF ACCELERATION — a body is inertial if and only if its motion creates no tension gradient, accelerating if and only if it creates a propagating tension wave. Every theory including general relativity distinguishes acceleration from inertial motion; that is dynamics, not Lorentz violation. T316 established the same independently, correcting "drag" to "added mass" precisely because velocity-dependent drag would give a detectable rest frame, and showing by d'Alembert's theorem that steady motion produces no force at all. (c) GLOBAL COSMIC ROTATION: T136 identifies Gödel rotation as rebound angular momentum, and the archive quantifies it from angular momentum conservation as ω ≈ 10⁻⁶⁸ rad/s, noting it is "consistent with the universe appearing non-rotating to current observational precision while nevertheless carrying a definite angular momentum imprinted by the rebound." Against the isotropy bound ω/H₀ < 10⁻⁹, THE FRAMEWORK PASSES BY FORTY-ONE ORDERS OF MAGNITUDE. RESULT 333.2 (two escapes fail). T185'S OWN ESCAPE FAILS: it suggests that "if the lattice tension λ varies significantly only at cosmological scales, the solar system constraint may not apply." But ω_BD is a PARAMETER OF THE THEORY, not a field value — it appears as a fixed constant in the field equation, as T185's own equation (4) has it, and is the same number everywhere. What varies is φ_BD, and Cassini constrains ω precisely by measuring how little φ_BD varies. AN APPEAL TO SCREENING ALSO FAILS: chameleon and symmetron models evade solar-system bounds by making the scalar heavy and short-ranged in dense regions, suppressing its response to matter. It is tempting to read T326's slaving as the same mechanism. IT IS NOT — SCREENING SUPPRESSES THE RESPONSE; SLAVING MERELY RELOCATES IT. If the field responds to matter, the multiplier responds through the field, unsuppressed. This analogy was proposed in the course of the present work and is withdrawn here. RESULT 333.3 (one route remains, independently motivated). The entire exposure enters through one line — T185's Theorem 4: φ_BD(x,t) = λ(x,t)/λ₀. THAT SINGLE IDENTIFICATION IMPORTS THE WHOLE BRANS-DICKE APPARATUS, its parameter, its γ, and the spacecraft's verdict. Remove it and Cassini says nothing about the framework. AND THERE IS INDEPENDENT REASON TO DOUBT IT, not arising from wanting to evade the bound: a Brans-Dicke scalar is DYNAMICAL BY DEFINITION, possessing its own wave equation □φ_BD = 8πT/(3+2ω) sourced by matter. T326's variation shows the multiplier has no such equation — it is determined pointwise as λ = Q/|G|², algebraically slaved. A SLAVED QUANTITY IS NOT A BRANS-DICKE SCALAR, so the identification may never have been licensed and the exposure it created may never have been real. This is NOT established here; what is established is that it is the only remaining route, that it was unavailable in April since T326 postdates T185 by four months, and that it costs whatever T185 derived through the identification. KILL-CONDITIONS: (i) if the scalar carries a mass, the PPN formalism does not apply conventionally and constant PPN parameters break down — massive Brans-Dicke theories evade the bound for scalar masses above ~2.5×10⁻²⁰ eV, and THE FRAMEWORK HAS NEVER ASSIGNED λ A MASS, so this route is open but unexplored; (ii) if the K-weight formula is revised and yields ω > 40,000, the exposure closes with the identification intact — T185 names this itself; (iii) if the Cassini measurement is superseded, the bound moves; (iv) if the framework's gravity is not scalar-tensor at all — which T302's no-go independently suggests, since a rotating substrate carries two fields rather than one — the PPN analysis must be redone with both sectors, and γ is not necessarily the parameter carrying the deviation. NOT CLAIMED: that the framework is refuted, since three of four exposures close and the survivor traces to one identification with independent reason to doubt it; that T185's Theorem 4 is wrong, argued as plausible not demonstrated; that the correct PPN parameters for a two-field substrate have been computed; that T326 dissolves the exposure, since Result 333.2 withdraws exactly that hope; or that any escape has been verified.