A concrete escape from the negative-energy requirement exists in the corpus, but only by leaving standard general relativity. Varieschi & Burstein (2012) recompute the Alcubierre metric in Conformal Gravity — a fourth-order theory whose field equations replace Einstein's — and find that with the Hartle shaping function the Eulerian energy density T^00 is non-negative, so no exotic matter is needed. The result is stated by its authors as strictly conditional: 'if CG is the correct extension of GR'. Two limitations the corpus makes visible: conformal gravity is a speculative alternative to GR, not the accepted theory; and the positivity demonstrated is of the Eulerian T^00, the same one-observer quantity Santiago-Schuster-Visser showed is necessary but not sufficient for the full weak energy condition, while the paper itself concedes the dominant energy condition is violated.
What it rests on
Attacks run against it
I attacked the headline sentence ("A concrete escape from the negative-energy requirement exists in the corpus, but only by leaving standard general relativity") with a dilemma built from counter-sources inside the corpus itself. Horn 1 (all-observer standard): if "escape" means the full WEC, the corpus's own GREEN machine-verified recomputation C-0028 (compute/bach_cg.py, report/bach_hsf_v1_minmax.json) shows the CG warp drive violates the WEC within Conformal Gravity at every speed tested — I verified from the report JSON that on the HSF v=1 grid a rapidity-2 observer at the bubble centre (x=0, rho=0) measures T_uu = -2183.2 while another measures +2183.2 (code units; Eulerian peak 919.3), both signs far above noise, so under either global sign convention there is no escape even granting CG. Horn 2 (Eulerian standard, the level C-0026's regime block fences 'no exotic matter' to): then "only by leaving standard GR" is attacked by Fell-Heisenberg S-0008, whose Eulerian-density positivity WITHIN standard GR the corpus certifies as proven (C-0021, GREEN, restricted ansatz), plus the subluminal positive-energy GR solutions S-0014/S-0015. Both horns failed to flip the truth value: Horn 1 is absorbed because the claim's regime block explicitly states "all-observer WEC is not demonstrated" and its invalid_outside names "any test of the full WEC over boosted observers, which the paper does not perform" — C-0028 settles exactly the question the claim disclaims, and the ledger itself records C-0028 as supports:[C-0026], with C-0028's computation CONFIRMING the claim's Eulerian sentence (one-signed to 1.4e-7 of scale, ASF control mixed-sign as the paper shows). Horn 2 is absorbed because the claim's regime is the Alcubierre metric family, and within GR the Alcubierre Eulerian density is non-positive for ANY shape function (C-0001, GREEN — Alcubierre's own theorem), so for that family "only by leaving GR" is rigorous; the within-GR Eulerian-positive papers all use different metric ansatze and are documented by the corpus as rebutted at the WEC level (S-0001 verbatim: they "merely assert the existence of one sub-class of timelike observers for which the energy density is positive. This is not enough to show that the WEC is satisfied"). I also independently re-verified all three evidence spans (E-0042 at char 13979, E-0043 at 637, E-0044 at 3135, sha256 3c2273b2... matching), confirmed the DEC-concession and author-conditionality sentences in the canonical fulltext, verified SSV's not-sufficient sentence verbatim in S-0001, and ran a live WebSearch for published comments/rebuttals of Varieschi-Burstein (none exist — consistent with the claim's posture and with C-0028 being the first independent check). A3 was probed (unqualified "T^00 is non-negative" vs the paper's own v_s <~ 2.5c cap in Fig. 2) and A4 (regime omits the velocity cap and the r=R distributional wall terms); both are real gaps but neither flips the regime-qualified conclusion, since the paper's flagship v=1 case is superluminal and inside the cap. Reproducible record: attack_c0026.py in my scratch dir runs all checks. || FINDINGS: (1) All three evidence quotes verify EXACT at their recorded offsets against canonical sha256 3c2273b201b168eb...; the "if CG is the correct extension of GR" conditionality and the DEC-violation concession are verbatim in S-0018. (2) C-0028 (GREEN, redteam-verified, supports:[C-0026]) reproduces the claim's Eulerian sentence: HSF v=1 grid one-signed, eulerian_min -919.35 vs opposite-sign residual +1.3e-4 (1.4e-7 of scale); ASF control mixed-sign (-2885/+872) as the paper shows. (3) The same C-0028 kills any unhedged reading: boosted observers see both signs at every speed (bubble centre +/-2183.2 = 2.4x Eulerian peak; v=3 worst -503995), but the claim's regime block pre-disclaims exactly this ("all-observer WEC is not demonstrated"). (4) Within-GR Eulerian positivity exists in the corpus (Fell-Heisenberg S-0008 per C-0021 GREEN)