The Closure on the Bench: A Spin Closure, Written in the Variables of a Condensate

One number does a great deal of quiet work in bouncing cosmologies built on Einstein–Cartan gravity: the closure s² = (ħcn)²/8, the squared spin density of an unpolarized fermion fluid, stated by Poplawski in 2010 and used wherever spin and torsion are asked to stop a collapse. My own earlier attack on that closure derived its shape — positive, quadratic in the density, with the signed spin averaging to zero while the square survives — and reduced its coefficient to one named prescription: build the continuum out of cells holding one particle each, and a factor of two in the gravitational action's dressing is still owed. This paper asks the next question, the cheapest one that can kill: can the closure be said at all in the variables of a laboratory bench — a spinor Bose–Einstein condensate, where spin density, filling, and coherence are things an experimenter sets with knobs and reads with cameras? The answer is yes, but not in the shape anyone would have guessed. The closure's bench image is not its famous number. It is a surface: for a condensate coarse-grained into cells holding N atoms each, at coherence factor P, the averaged squared spin density obeys ⟨s²⟩ / ((ħ/2)n)² = P² + (1 − P²)/N, verified here by Monte Carlo against the formula at better than one percent across the grid. The closure itself is one point of that surface — the point the one-particle-per-cell prescription picks out — and the extra factor of one half that turns the ensemble's 1/4 into the printed 1/8 belongs to gravity's action, for which no bench variable exists. I prove the companion negative in full: the bare number 1/8 has no bench fingerprint, because a whole curve of distinct bench states, at different fillings and coherences, sits on it. What a bench can test is the surface, and I pre-register the test with numbers for rubidium-87: at peak condensate density, a one-cubic-micron imaging voxel holds 260 atoms, and forced incoherence must suppress the voxel's spin-density square by exactly that factor of 260, or the ensemble mathematics underneath the closure is wrong in matter and not only in cosmology. Let me be plain about the fence around all of this: the bench tests the closure's mathematics — the averaging — and not any coupling of spin to spacetime. A pass would validate bookkeeping. It would say nothing about gravity, torsion in spacetime, or any bounce. Grade: ANALOG IMAGE EXISTS — AS A SURFACE, NOT A NUMBER; PRE-REGISTERED WITH KILL CONDITIONS. Keywords: spin closure; Einstein–Cartan; spinor Bose–Einstein condensate; coherence factor; filling; analog test; spin noise.