ITU and Quantum Gravity: A Single-Axiom View of AdS/CFT, Ryu-Takayanagi, ER=EPR, Page Curve, Loop Quantum Gravity, String Theory, and the 2026-2050 Experimental Roadmap

We apply the Information-Theoretic Unification (ITU) framework (Terada 2026, concept DOI 10.5281/zenodo.20109209; current latest Tier 0 v3.0 at 10.5281/zenodo.20200156) to quantum gravity. This is Tier 1 paper #17, opening Block A (Physics/Math deepening) and adding the K-skeleton hub to the ITU polytope, bringing it to 17 vertices. Tier 1 #17 connects to all 16 existing Tier 1 papers (degree 16), pairing with Smart Cities #16 (degree 15) as the polytope's deepest structural hub. Pass-1 progress: 118 of 220 phases (53.6%). Phase 111 establishes AdS/CFT and the Ryu-Takayanagi formula as the geometric realisation of the ITU axiom: under RT correspondence S(rho_A) = Area(gamma_A)/(4 G_N) reduces to delta S = delta <K_geom> with K_geom = 2 pi times the boost generator around the minimal surface. Numerical RT for AdS_3/CFT_2 disk reproduces S_CFT = (c/3) log(L/eps) with c = 1.5 at machine precision. K-state map: 16 Tier 1 to quantum gravity coupling averages 0.522 (top: QC 0.95, Energy 0.85, Communications 0.75). Phase 112 covers ER=EPR (Maldacena-Susskind 2013) and thermofield double. TFD reduces to thermal state with rel err 2.47e-15 over five beta values; entropy maps to bulk BH area S = A_h/(4 G_N). ITU first law dS = beta dE verified to mean rel err 6.91e-04. Hawking 1976 information paradox preview: naive Hawking vs Page-curve unitary, t_Page = t_evap/2. Phase 113 derives the Page curve from Haar-random states (Page 1993) with max |S_emp - S_Page| = 0.019 at d_total = 64, and shows the island formula (Penington 2019, Almheiri et al. 2019) as min ext of two saddles. ITU interpretation: island formula = K-channel competition between K_thermal and K_geom. Phase 114 resolves the AMPS firewall (2012) via ER=EPR identification: monogamy preserved without firewall. Planck-scale fundamentals: l_P = 1.616e-35 m, t_P = 5.391e-44 s, E_P = 1.221e+19 GeV. ITU UV cutoff f_ITU(x) = exp(-0.05 x^2) regulates trans-Planckian divergence (peak local frequency 31.64 to 1.92 bounded). Phase 115 develops loop quantum gravity (Rovelli-Smolin 1995). Area spectrum A(j) = 8 pi gamma l_P^2 sqrt(j(j+1)) gives A_min(j=1/2) = 5.169 l_P^2 at Immirzi gamma = 0.2375. Large-N puncture sampling converges to semiclassical mean within 0.1% at N = 10^5. ITU first law on discrete area jumps verified. gamma_ABCK = log 3 / (pi sqrt 2) = 0.247275. Phase 116 covers string theory and M-theory. Five superstrings (I, IIA, IIB, HO, HE in 10D) and M-theory (11D, Witten 1995) unified via T/S-duality (self-dual R = sqrt(alpha') = 1, self-dual g = 1). Strominger-Vafa (1996) S_micro = 2 pi sqrt(Q_1 Q_5 N) reproduces Bekenstein-Hawking A/(4 G_N) exactly. Three-route convergence on same S_BH at A = 1000 l_P^2: semiclassical 250.0, LQG 134.1, Strominger-Vafa 261.2 nats. Phase 117 surveys QG experiments. LIGO/Virgo 1e-22 strain detects 100+ BBH mergers; ITU echo h ~ 9e-62 below CE/ET 2030+ 1e-23 sensitivity by 39 orders. EHT M87* 39.7 uas / Sgr A* 50.7 uas; QG corrections l_P/R_s ~ 1e-45 out of reach for ngEHT 5 uas. LISA EMRI 2.77e5 cycles/4 yr; integrated correction 1.5e-39 below SNR. BMV (Bose-Marletto-Vedral 2017) phase Delta phi = G m^2 t/(hbar d) = 0.633 rad for m = 1e-14 kg, d = 100 um, t = 1 s exceeds detection threshold, making BMV the gold standard for direct quantum-gravity testing in 2030-2040. GRB 090510 (Fermi-LAT) excludes linear (n=1) Lorentz violation E_QG/E_P > 1.45. Phase 118 integrates all 7 prior phases under K_geom as the common modular-Hamiltonian backbone. The 17-vertex polytope: 17 vertices, 76 edges (60 existing + 16 new), <k> = 8.94. Tier 1 #17 degree 16 = K-skeleton hub (paired with #16 deg 15 URBAN hub). Ten falsifiable predictions for 2026-2050: BMV detection 2032 P=0.65, GW echo 2038 P=0.50, EHT M87* 5 uas 2030 P=0.85, LISA EMRI >=10 by 2038 P=0.75, CMB B-mode LiteBIRD 2034 P=0.60, quantum BH simulation 1M qubit 2040 P=0.55, atom interferometer 1e-13 by 2030 P=0.70, holographic complexity 2040 P=0.40, Sgr A* Kerr deviation <1% 2035 P=0.70, Lorentz E_QG > 10 E_P (n=2) 2032 P=0.55. Grand P_avg = 0.625 (>Tier 1 #1-#16 mean 0.59). Strong/Medium = 6/4. Central thesis: ITU axiom delta S(rho_A) = delta Tr[K_A^(0) rho_A] is the common backbone of all three quantum-gravity quantizations (AdS/CFT continuous bulk, LQG discrete spin networks, string D-brane microstates), with K_geom = modular Hamiltonian as the geometric K-state. The 17-vertex polytope demonstrates that quantum gravity occupies the deepest structural layer of ITU. BMV is the unique direct-experimental gateway (2030-2040). Pass-2 (Phase 221) will derive ITU-specific predictions: new HaPPY-type holographic codes, bulk reconstruction algorithms, quantum extremal surface numerics, replica-wormhole simulations, and ITU-specific BMV phase corrections distinguishable from Newtonian. Honest framing: this is a Pass-1 interpretive paper. Numerical results agree with established literature (Ryu-Takayanagi, Page, Strominger-Vafa, Rovelli-Smolin, BMV proposal). No new physical predictions beyond what the cited works already contain; the contribution is the unified ITU axiom narrative and the 17-vertex polytope structure. Includes 8 theory documents (theory_phase111-118.md), 8 Python numerical experiments, 8 figures (PNG), 8 JSON summaries. Total runtime ~12 seconds.