A bit of one, a bit of the other. It's fairly well-optimised. I think the big remaining optimisation will be to de-allocate memory less frequently — the program creates and destroys millions of structs representing types. 20% of the runtime is taken up iterating over and dropping collections of type-related structs.
Yes — that's after a ton of Rust-centric optimisation. About 25% of the runtime is currently consumed with deallocating memory (there are a lot of heavily-nested data structures getting cleaned up), so there's definitely some more work to be done to reduce cloning.
All engineering at sea costs more. Seabed anchoring is a thing, and a higher degree of wind offshore is a thing, but if you do the linear optimisation of the different cost benefit lines I suspect scaling up traditional fan style 3 blades just wins.
It's a "perfect is the enemy of good enough" thing. Better designs along one axis with a multi axis problem won't be best overall.
Interestingly, vertical designs have been proposed for Mars, because space for shipping them is at a serious premium, and the much lower atmospheric density reduces the stress considerably. (Hollywood depictions of destructive Martian sandstorms are quite exaggerated for effect.)
The pole in the center of the VAT generates a turbulent wake that the blades pass through once every rotation. This wake shakes the blade and causes extra stress. The blades themselves make a wake also, though not as much as the pole.
The horizontal turbines put the pole behind the blades. The pole does make a bow wave, but it is not as bad as the VAT wake. The pole/blade interaction is not as severe.
