Claims, non-claims, and falsification

We establish the substrate result (G_relaxed→0, B finite at z=2d, c_s²=B/ρ) and all four pillar statements; the entire residual is compressed to two named gates — global regularity and the infinite-Re flux mechanism — while A0 non-uniqueness, the companion’s speed-of-light reading of c_s²=B/ρ, and the adjacency φ_jam≈0.633 vs 2/π=0.6366 are held deliberately out of scope. [DERIVE]

This section is deliberately blunt. A foundational document is trustworthy in proportion to how clearly it bounds itself.

What we claim

The substrate is a fluid by arrangement: at the isostatic margin the relaxed shear modulus vanishes while the bulk modulus stays finite, leaving a single elastic (sound) speed $c_s^2=B/\rho$ (Section §3, [DERIVE]: local reproduction plus the companion's five-observable verification).
The continuum balance laws are exact consequences of a massive, interacting arrangement (Pillar I, [DERIVE]).
Given a rotational arrangement, the effective core length is fixed by an exact identity, and the cross-regime audit measuring departures is reproducible to machine precision (Pillar II, [DERIVE]).
A binding-versus-penalty arrangement selects a length with an exact one-half control exponent, recovered by nonlinear simulation and blind DNS (Pillar III, [DERIVE]).
Rearrangement events carry a nonnegative dissipation measure whose steady-state budget is exact and saturates at the injection rate as viscosity vanishes (Pillar IV, [DERIVE] for the budget). Those events are distributed as marginal-stability avalanches: in developed 3D turbulence the thresholded dissipation structures follow a scale-free size law with the pre-registered Lin–Wyart exponent $\tau\approx1.4$ – (measured at the higher resolution), confirming a genuine prediction of the substrate identification (\S§9, [DERIVE] for the category; the asymptotic exponent remains a [GATE]). On release into decay the dissipation coefficient $C_\eps$ follows the Vassilicos non-equilibrium law $\propto\Rey_\lambda^{-1}$ ( at the resolved ) and then relaxes onto the constant- $C_\eps$ plateau — the $\Pi_L$ event-RG fixed point, i.e. Taylor's zeroth law (\S§9, [DERIVE] for the category).
The substrate rules fix the universality class of the laminar–turbulent transition as directed percolation, in agreement with experiment (2016–2024) and verified on five exponents in this bundle (\S§3, [DERIVE] for the category).
Assembled, these say one thing: fluid form is configured.

What we explicitly do \emph{not} claim

The Navier–Stokes equations are derived, not overturned. They follow as the exact Newtonian closure of the arrangement (Pillar I) and are used throughout — this framework is the foundation beneath fluid dynamics, not a competitor to it.
Global regularity is compressed to one gate, not claimed. The configurational view recasts the Millennium question: coarse-grained fields are smooth at finite kernel width, and rearrangement events supply a physical regularization valve the field formulation lacks. What remains is a single, sharply stated zero-width limit — named as the lone open [GATE], not asserted.
We do not claim A0 is the unique microscopic substrate of fluids. A0 is a sufficient model: from it the pillars follow. Other substrates that reproduce the same coarse-grained identities are not excluded.
We do not claim the event channel is proven to carry the inertial flux in the infinite-Reynolds limit. We demonstrate the budget, the saturation, the exact 1D Burgers flux-carrying, the resolution-stable 2D concentration, and the resolution-converged 3D concentration (developed turbulence, skewness $\approx-0.5$ : forward energy flux on strain events, $\mathrm{corr}\approx0.80$ , converged across ); we report a negative control showing the mechanism is not automatic. What we do not claim is the asymptotic statement—that the support tends to measure zero as $\mathrm{Re}\to\infty$ —the genuine open Onsager/Duchon–Robert problem, which is the residual [GATE] of Pillar IV.
We do not import the companion's speed-of-light identification. The substrate sound speed $c_s^2=B/\rho$ of Section §3 is, in this document, an elastic-wave speed of the arrangement and nothing more. The companion's stronger reading (that this speed at full jamming is the invariant speed) rests on its dimensional anchor and is neither used nor claimed here.
We do not claim identities from numeric adjacency. In particular, the measured jamming fraction $\phi_{\rm jam}\approx0.633$ and the attractor constant $2/\pi=0.6366$ are adjacent numerals, and no identity, cause, or evidence is claimed from that adjacency anywhere in this document (the companion states the same rule for itself).
We do not reinterpret any measurement. Every phenomenon is taken as recorded by standard fluid mechanics (Principle §2).

Falsification criteria (per pillar)

Each pillar names a concrete way it could be wrong. Gates are listed as gates precisely so they can be attacked.
Pillar	Falsified if \dots
0	a packing prepared and accepted under the stated substrate protocol shows a relaxed shear modulus that fails to vanish toward (bootstrap confidence interval excluding as grows), or a bulk modulus that fails to stay finite there; or the contact-breaking shear reserve fails to fall with pressure.
I	a declared regime passes the linearity/isotropy/short-memory tests yet the stress is measurably non-Newtonian, or the budget identity $\dot E=-\eps_\nu$ fails at fixed resolution beyond discretization error.
II	the RCCI identity is violated for a correctly computed mask/partition (algebraically impossible, hence a code/measurement check), or the regime/method ordering of the audit reverses under reasonable symmetric error metrics.
III	the selected length fails to scale as $(\sigma/\varepsilon)^{1/2}$ (exponent not ) in a flow that genuinely realizes the binding/penalty dispersion, or the forward/shuffle slope test stops distinguishing real selection from re-pairing.
IV	a forced steady flow with vanishing viscosity shows total dissipation falling to zero with $\nu$ (no anomaly), or event activity demonstrably fails to concentrate at the under-resolved increments while the anomaly persists; or the thresholded dissipation structures of developed turbulence are distributed with an intrinsic scale (an exponential rather than a power law), or with a size exponent that does not approach the marginal-stability band as resolution grows; or a forced field released into decay shows the dissipation coefficient $C_\eps$ staying constant through the non-equilibrium transient (no $\propto\Rey_\lambda^{-1}$ flow), or failing to relax onto a plateau as the decay becomes self-similar.
T	the laminar–turbulent transition of a subcritical shear flow is conclusively shown to lie outside the directed-percolation class, or any of the five lattice gates of \S§3 fails on rerun.

Open gates and the research program

The live work is concentrated in a small number of gates. Two now sit at the substrate (Section §3): exact marginality under drive—whether driving takes exactly to rather than pinning near it (gate G-SOC; the companion's pinning evidence is $g^\star/g_0\approx1.1$ – 1.3 )—and the finite-rate divergence of the viscosity, so far measured only through its static and quasistatic proxies. Within the pillars, the first is the Newtonian-closure gate of Pillar I—now located on the $\Delta z$ axis of Section §3: measure $\mu$ from a concrete arrangement by Green–Kubo and confirm the regime conditions. The second is the flux-carrying gate of Pillar IV—and here the status has advanced substantially. In one dimension it is closed: in Burgers turbulence the dissipation equals the shock defect $(\Delta u)^3/12$ , $\nu$ -independently, with the Duchon–Robert inter-scale flux concentrated at the event (Section §9). In two dimensions the forward enstrophy flux concentrates on the intense vorticity-gradient events ( $\approx60\%$ carried by the top $20\%$ of the area, $\approx12$ – $16\%$ of the area carrying half), co-locates with the dissipation density, and is stable across N=192,256,320 . What remains is the infinite-Reynolds asymptotic question—whether the flux support tends to a measure-zero set as $\mathrm{Re}\to\infty$ . The finite-Reynolds mechanism is now established in all three dimensions: exact in 1D, resolution-stable in 2D, and resolution-converged in developed 3D turbulence (ns3d.py: forward energy flux on strain events, top- $20\%\approx60$ – $66\%$ , $\mathrm{corr}\approx0.80$ , converged across N=64,80,96,128 ). Closing the asymptotic statement—the genuine Onsager/Duchon–Robert problem, beyond any finite simulation—would complete the upgrade of Claim §9 from a budget statement to a fully proved mechanism. The remaining frontier is the dimensional anchoring of gate G-S: the selection law and the JFM exponent are dimensionless, so reaching an absolute quantum length requires the $D=2\pi\lambda/A$ mapping, which is referenced rather than derived. All three—the Newtonian closure, the flux asymptotic, and the G-S anchoring—are falsifiable and scoped in Table §14.