I'm still unsure how you think this contradicts the claim that plants built together, in series, were cheaper than building smaller numbers of plants. Yes, those plants took 5-10 years to build. And? This doesn't change the fact that they were built for ~$1.5B per GW of capacity. The last of these plants were finished as construction started on later plants that would come online during the 1980s or 90s, and those later plants were much more expensive. But there were also fewer of them which leads to less economy of scale.

Let's put this in simpler terms:

* Lots of plants that started construction in the 1960s-1970s time frame -> lower cost per GW

* Smaller number of plants started in the 1970s and beyond -> higher cost per GW

Thus, a strong relationship between a greater number of plants build built and lower cost per plant. Construction time frame doesn't matter much. If 100 plants are started in 1965 and finished in 1980 and 10 plants are started in 1975 and finished in 1990, the 5 years of overlap doesn't change the fact that the former is going to have a much better economy of scale than the latter.