s1208 — S369.19: octant invariant census on the assembled lepton pair (S370)
  served: sin²θ12 0.304442 · sin²θ13 0.021447 · δ_CP 199.0031° (cos δ -0.9455) · octants sin²θ23 ∈ {7/16, 9/16} · 25 necklace words, length 2..6
  CP-EVEN (Re Tr w): 25 of 25 words SEPARATE the octant · 0 blind · 0 unresolved
     XY      Re +4.4144640802e-03 → +3.4723693526e-03  rel Δ 2.134e-01   SEPARATES
     XXY     Re +1.3899800232e-02 → +1.0914637180e-02  rel Δ 2.148e-01   SEPARATES
     XYY     Re +1.0896575347e-05 → +8.4927151920e-06  rel Δ 2.206e-01   SEPARATES
     XXXY    Re +4.3887911390e-02 → +3.4462200668e-02  rel Δ 2.148e-01   SEPARATES
     XXYY    Re +3.4311646542e-05 → +2.6694668860e-05  rel Δ 2.220e-01   SEPARATES
     XYXY    Re +1.9478620565e-05 → +1.2048476380e-05  rel Δ 3.815e-01   SEPARATES
     XYYY    Re +2.7231114671e-08 → +2.1217868102e-08  rel Δ 2.208e-01   SEPARATES
     XXXXY   Re +1.3857520749e-01 → +1.0881371097e-01  rel Δ 2.148e-01   SEPARATES
     … lowest separating word: XY (degree 2)
  CP-ODD  (Im Tr w): nonzero on 2 words, all of length 6: ['XXYXYY', 'XXYYXY'] · of those, separating: 0
  T1c ΔRe Tr(H_e H_ν) numeric -9.420947276871e-04 · from Σ m_α² m_i² Δ|U_αi|² -9.420947276871e-04 · e-row Δ|U_ei|² max 0.0e+00
  T4 max_{θ23, δ grid} min_{D1,D2} |U(90°−θ23, δ) − D1 P_μτ U(θ23, δ+π) D2| = 1.1e-16
  T5 m_μ = m_τ, δ = 90°: all 25 words blind: True · m_μ = m_τ, δ served (cos δ -0.946): 0 CP-even words separate (lowest None)
  T5c m_μ ≠ m_τ, δ = 90° (cos δ = 0): 25 CP-even words separate (lowest XY)
  VERDICTS: octant visible: YES — CP-even, lowest degree 2 (Re Tr H_e H_nu) · CP-odd blind: YES — every CP-odd invariant to degree 6 · mechanism: flip = mu<->tau relabel o (delta -> delta+pi); visible only through m_mu^2 - m_tau^2
  CHECKS: {"argument": 3, "computed": 12, "failed": 0, "receipt": 1, "symbolic": 0}
    PASS R1 s1204 served rows (the assembled object, same receipt as s1205 R1)
    PASS T1 NOT A NO-GO: at least one CP-even invariant separates the octant (rel Δ > 1e-9)
    PASS T1b the lowest separating invariant is degree 2: Re Tr(H_e H_ν)
    PASS T1c the degree-2 difference equals Σ m_α² m_i² Δ|U_αi|² (rel 1e-10), and the e-row moduli do not move (the octant live
    PASS T2 every CP-odd invariant is octant-BLIND: the nonzero Im Tr w (length 6 only) agree at both octants to 1e-12
    PASS T2b no word shorter than 6 has a CP-odd part (|Im|/|Tr| < 1e-12)
    PASS T3 CONTROL: one-letter traces (U-independent) are unchanged (rel < 1e-12)
    PASS T3b no word is UNRESOLVED (every rel Δ is < 1e-12 or > 1e-9)
    PASS T4 octant flip ≡ μ↔τ relabel ∘ (δ → δ+π), up to diagonal signs (residual < 1e-14 on a 4×4 grid)
    PASS T5 FIRING CONTROL: m_μ = m_τ and cos δ = 0 ⇒ EVERY invariant is octant-blind (the census can report blindness)
    PASS T5b (rewritten to the finding; E5 second half falsified) m_μ = m_τ at the served δ, cos δ ≠ 0 ⇒ STILL every invariant 
    PASS T5c the m_μ² − m_τ² channel ALONE suffices: nondegenerate masses at cos δ = 0 ⇒ Re Tr(H_e H_ν) separates
    PASS T6 second mass set: Re Tr(XY) separates and every nonzero CP-odd part is blind (same verdicts)
  RESULT_HASH a057865a49e4957c
