RQ-A U.2 / U.3 (CLAIMS 3a8a0c64), K = 4, 5, 6
PASS K=4: g((K-1)s(q_K)) = g(K s(q_K)) (|diff| < 1e-70)
   C(q_4) = 0.85334356849604864915   eta_4 = 0.0044203164646581473302
PASS K=4: C(q_K) = 0.8533435684... (truncated)
PASS K=4: eta_K = 0.0044203164... (truncated)
PASS K=4: margin 0.00443 >= eta_K
PASS K=4: sqrt(2/e) - 0.00443 <= C(q_K)
PASS K=4: example C = 0.8533435 <= C(q_K)
PASS K=5: g((K-1)s(q_K)) = g(K s(q_K)) (|diff| < 1e-70)
   C(q_5) = 0.85510045547081924572   eta_5 = 0.0026634294898875507617
PASS K=5: C(q_K) = 0.8551004554... (truncated)
PASS K=5: eta_K = 0.0026634294... (truncated)
PASS K=5: margin 0.00267 >= eta_K
PASS K=5: sqrt(2/e) - 0.00267 <= C(q_K)
PASS K=5: example C = 0.855100 <= C(q_K)
PASS K=6: g((K-1)s(q_K)) = g(K s(q_K)) (|diff| < 1e-70)
   C(q_6) = 0.85598448917381538818   eta_6 = 0.0017793957868914082978
PASS K=6: C(q_K) = 0.8559844891... (truncated)
PASS K=6: eta_K = 0.0017793957... (truncated)
PASS K=6: margin 0.00178 >= eta_K
PASS K=6: sqrt(2/e) - 0.00178 <= C(q_K)
PASS K=6: example C = 0.855984 <= C(q_K)
PASS U.3: 0.853344 > C(q_4) (the old text rounded up, so the fix was needed)
PASS U.3: every decimal with 7 places that is <= C(q_4) is <= 0.8533435
RQ-B C19 remark (CLAIMS e7973f59): theta <= 1/sqrt(4 s(q4)) = 1.3183436
   1/sqrt(4 s(q4)) = 1.3183435623505367029
PASS |1/sqrt(4 s(q4)) - 1.3183436| < 5e-8 (correctly rounded to 7 places)
PASS 1.318 < 1/sqrt(4 s(q4)) (the old decimal overstated the necessary condition)
   1/sqrt(4 s(q4)) - 1.318 = 0.000343562
PASS the overstatement was 3.4e-4 to two digits

ALL CHECKS PASSED
