Uses the (then still-pending on appeal) Wagner/Sancho litigation as the vehicle to ask what a court should do, as a matter of evidence and procedure, when asked to enjoin an experiment on the grounds it might destroy the Earth — a case where standard tools (expert deference, cost-benefit/law-and-economics analysis, the preponderance standard) arguably break down because the harm is unbounded, the probability is contested at many orders of magnitude by non-neutral experts, and irreversibility makes ordinary injunction balancing degenerate. Johnson proposes analytical tools for meaningful judicial review of extremely technical low-probability/catastrophic-stakes claims without simply asking judges to referee the physics. Concludes the physics community’s self-policing (LSAG) plus procedural dismissal, as actually occurred, was defensible but leaves the “who checks the checkers” question open for future cases with less scientific consensus. relevance_note: The most direct treatment of whether courts/legal process can or should independently verify a physics safety case — a distinct “hinge” from the purely scientific ones.