This page records empirical consequences only when the program says something that could be wrong. A consistency check is not listed here. Reproducing a known invariant, a known scaling law, or a known geometric structure may validate the machinery, but it is not a prediction unless the result could have disagreed with an external comparator.
At present there is one public prediction: a charge-sector exclusion. Numerical confinement benchmarks from Appendix B are not listed here, because they remain benchmark/interface work rather than a theorem-certified empirical prediction.
Statement. Assume the closed electromagnetic phase sector constructed in Chapter 18 of Closed Systems from Comparison Completeness is physically realized, and normalize the primitive charge unit by the observed electron charge magnitude e. Then every free/asymptotic vacuum electromagnetic charge lies in the denominator-3 lattice:
q/e ∈ (1/3)ℤ.
In particular, there is no free/asymptotic fundamental particle with true vacuum electromagnetic charge
0 < |q| < e/3.
This is a forbidden-region prediction. It does not predict a new particle. It predicts that confirmed free/asymptotic particles with true vacuum electromagnetic charge outside the denominator-3 lattice do not exist; in particular, no true millicharged state exists in the band 0 < |q| < e/3.
The technical source is Chapter 18 of the monograph. The phase sector is first made discrete by closed-system admissibility. The charge assignment is then a group homomorphism from the global phase-sector group to the real numbers. If the global phase sector is generated by a primitive phase sector p0, the charge image is the rank-1 lattice q0ℤ. Thus all primitive fundamental charges are integer multiples of a primitive charge unit q0.
Fractional observable charge is not introduced at the primitive phase level. It can arise only by observable descent of composite sectors. The first nontrivial composite denominator forced by the transport spine is 3, giving the minimal effective observable unit q0/3. With q0 calibrated to the electron charge magnitude e, the physically realized charge spectrum is constrained to the denominator-3 lattice.
Technical source: CSM, §§18.9–18.11, especially Theorem 18.10.4, Corollary 18.10.5, Remark 18.10.6, Theorem 18.11.5, Theorem 18.11.6, and Corollary 18.11.8.
Millicharged-particle searches are the cleanest current interface because they look for particles whose electric charge is much smaller than the electron charge. The closure prediction is not a cross-section, event-rate, or flux calculation. It is stronger and narrower: if the charge is a true free/asymptotic electromagnetic phase weight, then it must lie in the denominator-3 lattice.
For example, searches in the range
|q|/e ∈ [10^-3, 2×10^-2]
sit entirely inside the forbidden band 0 < |q| < e/3. Under Prediction P1, such searches should remain null for true free/asymptotic electromagnetic particles, apart from background, detector artifacts, or states whose apparent response is not a genuine vacuum electromagnetic phase weight.
Experimental context: CERN’s FORMOSA/FASER discussion of millicharged searches, and the LUX-ZEPLIN atmospheric millicharged-particle search, which reports sensitivity to charges between 0.001 and 0.02 of e and no significant excess in the reported exposure.
A confirmed free/asymptotic particle with true vacuum electromagnetic charge
q/e ∉ (1/3)ℤ
would falsify the closure electromagnetic sector as a universal physical charge theory.
In particular, a confirmed stable/free particle with charge such as
10^-3 e, 10^-2 e, 0.1 e, e/6, or e/10
would be a direct failure of Prediction P1, provided the charge is established as a genuine vacuum electromagnetic phase weight rather than an effective medium response or detector reconstruction artifact.
This page does not claim a reactor prediction, a fusion-confinement number, a dark-matter abundance, a production cross-section, or an event rate. It also does not claim that every null millicharged-particle search confirms closure. Null searches are consistent with Prediction P1; a confirmed charge outside the denominator-3 lattice is what would decide against it.
The point is the falsifiable exclusion: closure permits integer primitive charges and denominator-3 composite observable descent, but it excludes free/asymptotic fundamental millicharges.