They are reliable only on average and on long time scales. On one particularly abnormal weather day, do you just say: "sorry, we only have 50% power today".
PV is so cheap that if it was only 50%, one would just install twice as much.
I think the actual number may be 90% (daytime; obviously more like 100% at night, for which you have some storage).
But at this scale you want multiple types of power anyway just to avoid correlated failures, cf. “all of our coal plants have had to shut down at the same time because the coal miners organised a national strike”.
Question then is: is the cheapest solution to this already green, and if not, what do we need to invent such that the cheapest solution is also green?
From what I’ve seen, batteries at this scale are currently on par with nuclear for cost, but that’s the easiest hurdle to pass given how expensive nuclear is. Global grid is theoretically fine, and the purchase cost isn’t ridiculous, but I don’t know anything about geopolitical issues so it could be anything from “a way to bring the world together” to “laughably naive”, and I don’t know anything about ongoing maintenance costs. Hydrogen is cheap, I don’t know the metallurgy issues. Hydro is great, but has its own issues.
