Part of this came down to the logic of pursuing economies of scale. It costs a lot to build a 750MW plant, but not twice as much to build a 1500MW plant. So the

Part of this came down to the logic of pursuing economies of scale. It costs a lot to build a 750MW plant, but not twice as much to build a 1500MW plant. So the utilities leaned towards bigger plants (for everything, not just nuclear). But bigger plants take longer, overall. That means you need more capacity in the development pipeline, because you're forecasting that the market is growing. Soon you're needing to have finance and regulatory movement for multiple gigawatts of capacity at all times to keep up with what you think is coming. If you then find you've overshot demand, you are in an ugly position. Utilities that did the "dumb" thing and prioritised smaller plants, which are ostensibly more expensive per kwh, tended to do better because they didn't need to tie up so much capital at any given time. Bringing a gigawatt online in five years as a single plant is very different from bringing 200MW a year online. Especially if it turns out in the third year that 600MW was as much as the market wanted to buy.