The Missing Fifteen Ten-Thousandths: First-Row CKM Unitarity, Priced

Unitarity is not a fit. Inside the Standard Model, the first row of the quark mixing matrix must satisfy |V_ud|² + |V_us|² + |V_ub|² = 1 exactly — it is the statement that probability is conserved when an up quark decays. The measured entries do not sum to one. The current global fit gives |V_ud| = 0.97379(25) and |V_us| = 0.22405(35), and their row sums to Δ_CKM = −0.00151(53): fifteen ten-thousandths of probability are missing, at 2.8 sigma. And the size of the hole depends on a clock: the fit's neutron-decay input uses the bottle measurement of the neutron lifetime, 877.75(36) s. Feed it the beam measurement instead — 887.7(2.2) s, the ten-second defect priced in my paper *One Neutron, Two Clocks* — and the deficit deepens to −0.00234(62), 3.8 sigma. This paper prices the missing fifteen ten-thousandths: the three places they can hide (a radiative correction, a nuclear-structure correction, or new physics), the internal quarrel among the |V_us| determinations that shares the same stage, and the structural lesson — a unitarity test is a test of every correction spent on the way to it, and the two largest corrections in this test are each one calculation deep. Grade: DEFICIT STANDS — PRICED; ITS SIZE RUNS ON A NEUTRON CLOCK. This paper stands outside my work on the vacuum catastrophe; it draws nothing from that work and repairs nothing in it.