EEML

The previously maintained general dominance prohibition H_D remains visible as a historical strong hypothesis. The current D1 is a conditional process test: selection, inflow and reversibility are measured separately; the outcome alone does not classify the dynamics. Four constructive negative controls and the coefficient comparison with entanglement equilibrium have been newly carried out. The latter does not fit the positive K3 cost family. H_R has been registered separately as an information-conservation hypothesis; the Physarum design has been filled in, with explicit data requirements. No new measurement of nature or external preregistration has been carried out.

Added: the limit of canonical coupling selection, the relation of the K3 costs to quadratic gravity, the distinction between lower bound and coercivity, three family-counting routes, the 6D tensor condition and a register of test proposals. The K3 minimum theorem remains conditionally valid. A positive cost minimum does not prove a unique scale, Lorentzian unitarity or independent physical selection. External reports that calculations have been rerun count as reviews; their execution has not been separately inspected here.

The criticality line, inputs and error components have now been supplied and recalculated. The item is recorded as a proposal from the review of 28-09-2026, without external preregistration. The proxy gives \(M_t^{\rm crit}=170.952\) GeV and a central distance of \(1.568\) GeV. Combining the stated error budget in quadrature gives \(\delta_{\rm crit}\simeq1.41\text{–}2.24\), depending on the assumed additional mass-interpretation uncertainty. This is a sensitivity scenario, not a validated discovery significance. The 58 real K3 metric parameters including scale, and the earlier corrections, are retained.

Part VI.E now contains an explicit attempt: the existing K3 bound leaves the electromagnetic normalization free; an additional two-mode model can, however, fix a shape parameter by letting two contributions balance. The calculation and a chosen relaxation dynamics have been checked. The value 1/137 has not been derived. Flux choice, physical embedding, normalization and quantum corrections remain open.

Closure is now assessed per predicted quantity, distinguishing between dimension, number of solutions and number of distinct outcomes. The protocol contains a separate search for simultaneous saturation of bounds. The α attempt has been extended to remaining K3 freedoms, massless gauge fields and an explicit IIB flux family. In it, couplings can be fixed within a chosen sector, but the sectors examined do not yield a unique electromagnetic constant.

The supplied proposal for three adaptive routes has been included as the next experimental candidate, after checking the blockage law, measurable memory definition, parameter determination and reference models. The decrease theorem remains valid under the blockage specified below. Decrease alone, however, does not distinguish the process law, and a quiet baseline does not determine its rates. P1 is a checked design proposal, not a performed measurement of nature or a complete preregistration.