LFCT Validation Suite: Quantitative Predictions vs Observational Data (Data Runs E, F, G, H)

LFCT Validation Suite — Data Runs D, E, F, G, H Five independent validation runs for Light Frame Cadence Theory (LFCT), testing 59 quantitative predictions against published observational data. All predictions derive from axioms A1–A4 using the same structural constants (κ_TD = 1, κ_TS = π², κ_TR = 5/2, ε = 1/π², C₀ = 1/c, g_s = 5/7) with zero free parameters once the upstream structural scales are fixed. Each run is a self-contained Python script (Python 3.8+, NumPy) that prints a full scorecard and exports results to JSON. Run D — What LFCT Explains That Standard Cosmology Cannot (14 tests, Score 0.93) Explanatory-power validation across open problems: dark matter replacement via TS enhancement (SPARC RAR), dark energy as TR routing geometry, Hubble tension from two-fabric wrapping mismatch (0.2%), MOND acceleration scale (1%), CMB temperature, acoustic scale, odd/even peak asymmetry, damping tail, and low-ℓ anomalies. Run E — LFCT's Numbers vs the Sky (15 tests, Score 0.98) Number-vs-number validation of the LFIS-24 derivation chain. CMB first-peak position (0.9%), peak height ratio, damping scale, power spectrum amplitude (0.8%), Hubble constant ratio (0.2%), energy-density fractions, galaxy rotation curves, and nine-regime census. Run F — LFIS-27 Predictions (8 tests, Score 1.00) CMB architecture predictions: first-peak oscillation amplitude 28/5 (0.1%), peak spacing ratio (0.7%), electroweak crossover temperature (0.4%), Sachs–Wolfe plateau shape (2.1% RMS), and peak-2 seam correction. Run G — Slip, Polarisation, and Geometry (7 tests, Score 0.94) Physical Derivations predictions: gravitational slip η − 1 = −1/π³ (0.4σ), ISW enhancement ×1.066 (0.3σ), EE/TT normalisation (1.5%), B-mode structural suppression BB/EE = 2/(3π⁴), and CMB last-scattering radius (2.5%). Run H — Nine-Correction CMB TT Production Model (15 tests, PASS) Canonical validation of the LFIS-29 cadence star model. Reproduces the full CMB TT power spectrum (ℓ = 2–2508) — eight peak heights and seven trough floors — using additive density + velocity channel architecture derived from the Pascal limaçon B(φ) = π² cos φ + 1. Nine corrections, each with amplitude fixed by structural algebra. Result: HRMS = 0.94%, TrRMS = 0.94%, combined Score = 1.88%, all 15 features within ±2%. Combined results across Runs D–G: 27 PASS, 3 CONSISTENT, 0 FAIL. Run H validates the production model as a composite test at sub-2% precision. This archive contains two zip files: the original LFCT_2026_DataRuns_D-G.zip and the new LFCT_2026_DataRunH.zip, plus the shared README. Observational comparison data drawn from Planck 2018 (VI, V, XXI), SH0ES (Riess et al. 2022), McGaugh–Lelli–Schombert 2016 (SPARC RAR), BICEP/Keck 2021, DES Y3, and CAMB reference spectra. Requires: Python 3.8+, NumPy. Framework versions: LFIS-24 v2.0.3, LFIS-25 v1.1.8, LFIS-26 v1.3.1, LFIS-27 v1.3.2, LFIS-28 v1.3.1, LFIS-29 v1.0.2, Core CS v3.0.1, Core MF v3.0.1, Core PD v2.0.2.