[JonathanMace]

Reviewer B Round 2 — targeted surgical fixes

Fix 1: sigma_0 > 0 → sigma_3(0) = 0 (CRITICAL)

• Replaced sigma_min(delta) = sigma_0 + lambda_1delta^2 with correct sigma_3(eps) = lambda_1eps^2 + O(eps^4)
• sigma_3(0) = 0 at exact sphere by symmetry (geometric Jacobian columns proportional)
• Ritz intercept is discretisation artefact: grows with basis size (0.006, 0.011, 0.018 for 3/5/7 modes)
• Updated abstract, results, conclusion consistently

Fix 2: sigma_1, sigma_2, sigma_3 table

• Already present from R1 (verified complete with all singular values)

Fix 3: Precise prolate branch definition

• Added explicit formula a = a_0*(c_0/c)^{1/2} with c in [c_0, 1.5*a_0 = 0.27 m]
• Volume V = (4pi/3)a^2c and h held fixed
• c/a from 1.0 to ~1.5, all material parameters at canonical values

Fix 4: Computational methodology

• 5-mode Ritz with Legendre-polynomial trial functions in oblate spheroidal coords
• Central finite differences with Delta_theta/theta = 1e-4
• SVD yields sigma_1 >= sigma_2 >= sigma_3 and kappa = sigma_1/sigma_3

Fix 5: Manuscript-history rhetoric

• No remaining rhetoric found (cleaned in R1, verified with grep)

Fix 6: Precise prolate language

• kappa_prolate in range 156-244, no systematic improvement with asphericity
• Updated results and conclusion

All 487 tests pass. LaTeX compiles cleanly (6 pages).