Everything this model claims about the world that general relativity, collapse models, and Yukawa modifications do not. Each entry names its status, its registration, and what would kill it. A theory is worth exactly the column below.
A mass superposition at rest decoheres not at all; modulated at frequency ω, it decoheres along the exact Lorentzian (dT₀₀/dt)²/(γ²+ω²), knee at the drainage rate. GR: nothing at any frequency. Collapse models: a floor already present at rest. One frequency sweep distinguishes all three. Full experiment page →
The weak-field PPN parameter is fixed by one exponent: how the vacuum's audit rate scales with local stiffness. On this reading, Cassini's γ = 1 ± 2.3×10⁻⁵ is a six-digit measurement that the ledger's clock does not care what it prices: |a| < 5.8×10⁻⁶. Prediction: any future γ deviation must come with a knee that sags across potentials — the two are one parameter. Valid in the strong-monitoring regime, under six named assumptions.
The monogamy budget is finite: the vacuum cannot be drained below zero. Measured in the toy: at extreme consumption the effective coupling falls 7.6× — gravity weakens per unit mass² where the vacuum is exhausted — with an unexpected strengthening bump (+19%) just before saturation, flagged exploratory. Structural consequence: potentials are bounded by budget depth. This class of theories makes no singularities.
In one dimension the conserved flux is constant, so saturation is global: past a critical total consumption the sum rule has no solution at all — the universe cannot bookkeep more mass than its budget. Found as a "failure" of a registered gate; recorded as the structural prediction it actually is.
In d ≥ 2 the flux dilutes with distance, so saturation localizes: a super-critical mass grows a contiguous core of maximally-drained vacuum where no further drainage is possible — a horizon analog produced by accounting exhaustion, not geometry. The registered target: the core-radius law r(M), the model's Schwarzschild analog. Habitat: horizonless-object and echo phenomenology.
Inertia here is the gap-weighted fidelity metric of the vacuum: it depends on the vacuum's spectrum. Heat the vacuum, move a boundary, put a horizon behind it — the inertia shifts while the gravitational charge does not. A quantitative equivalence-principle-violation channel, δ(mi/mg) as a function of vacuum temperature: Mach's principle finally making a number instead of a speech. Deferred until the thermal Gaussian fidelity machinery is built — we do not guess formulas anymore.
The idle cost of the vacuum monitoring itself is a vacuum energy: Λ ~ ħγ · (idle information density). If the observed Λ fixes the drainage rate, the decoherence knee is pinned near cosmological rates and no knee exists in any laboratory band — which still discriminates: drainage-with-cosmological-knee predicts clean superpositions where collapse models predict a floor.
Below the substrate scale, the measured dressing of Newton's law oscillates in sign with a ~1/r envelope — no Yukawa modification does that. A standing falsifier: the day sub-millimeter (or any sub-substrate) experiments see a smooth monotone deviation, this class dies; the day they see an alternating one, nothing else predicted it.