Being written
This post is still on the drawing board.
The experiment has been run; the write-up has not. Rather than publish a half-argument, the article waits until the method, the figures and the failures can be set out properly — the failures are half the value.
What this post will cover
The starting point is an exercise carried out at the Politecnico di Milano: blades whose geometry follows Bézier curves, a solar coefficient computed for each variant, and the best one kept. The limit was the method — a random draw. The aim is to replace that trial and error with a convergence.