|Why to the cube?|
Necessarily greater than squared. Determined by experience. You get no economies of scale, bur rather the opposite. For example, as the array gets larger wind sheer becomes more significant so further design and construction are needed. Weather factors are killers in many ways. How do you manage a frost whose water freezes overnight? Install heaters?
I would expect < O(n^2)
You mean < O(n^k), where 2 < k <= 3.
We did discover that wind is the worst
Electro-mechanical. Mechanical devices exposed to weather break down chronically. Down time when the wind isn't blowing or the machine isn't working. Differential output across a field of turbines. that's no problem? Just rig some capacitors and resistors and you're done? Nope. Each rig needs its own service lines. In comparison to oil derricks you can't use the land around the rigs. They tried to do otherwise at a location not far from the Livermore Rad lab. They got fired beef. So they moved the farm to the CA Suisun Straits. Now it has problems with moisture coming from the delta. The problems are endless but fortunately, like all alternatives, its got a hefty subsidy. The fact that it's seriously uneconomical is swept under the rug based on the CO2 displacement factor.
Wind is as cheap as coal.
That is a lie. Someone is deceiving you. There's a whole lot of pseudos deceiving the German government and the people don't know it.
In fact, ALL of Germany's solar output is uneconomic. The pseudo scientists who argue otherwise, are bald faced liars. Once Germany gets rid of the EU you'll be finding out what a total fraud has been perpetrated. It's costing Germany something like 5B euros/yr to subsidize solar alone.
Maybe you can explain why both Spain and Germany are cutting back their government subsidies for solar. If it's so desirable, would they do it to save on expenditures?