So the thing about the earth is that it's big. If you have a big enough grid, it will be sunny and windy over a fairly consistent portion of that area.
Looking at maps for power prices with large hydropower dams you can easily see how most countries grids dont transfer more than around 1000km, and often quite a bit shorter. In practice this is the point where it becomes more cost effective to build new powerplants, than bigger infrastructure for transfer.
We could build grids designed for this, but never across oceans, and we are talking about an absolutely gargantuan investment, and its generally not considered as a solution for the unreliability of solar and wind. Its also why no one ever suggests to use the big empty oceans for truly large scale wind or solar etc, its always places very close to land.
Its also worth pointing out that superconductors cannot solve this, even if price and the gargantuan cooling energy requirements are ignored. This is because while superconductors transfer power without loss, they can only do so up to a limited effect, as the power interferes with the superconducting ability after that, essentially breaking it.
That sounds bloody swell. I'd pay for that in a heart beat.
That's an incredibly low amount. Taking Berlin as an example, even 5000km lets you reach plenty of hot, sunny desert areas where any losses will be easily offset by vastly increased solar efficiency. Or you could spread out your wind energy sources to get more consistent coverage.
In fact that's pretty much what Singapore is doing: building a solar farm 5000km away in Australia.
1000km of ocean reduce transmission to about 0.1%^^
Atmospheric depressions are bigger than than, and the distance for solar are much bigger as well (other than cloudy, there's day and night).
Specifically, https://en.wikipedia.org/wiki/Cost_of_electricity_by_source#... puts the cost of wind/solar at roughly 1/3rd the cost of nuclear, so it's more (cost) efficient to spend as much on transmission as you do on generation and build a really well connected massive grid than to use nuclear.
Also there is wind energy which doesn't care about whether the sun shines or not (actually it generates more energy when the sun isn't shining: https://www.scientificamerican.com/article/wind-power-turbin...).
Transmitting power also consumes energy - you cannot have a electric wire across pacific to transmit power from US to EU without transformers - electric power line is not a fiber optic that transmit just signal.
I also doubt we have enough rare earth minerals to make all those magnets for wind turbine and and electric car enough for whole planet. Remember also you have to decomission those windturbine after some years which is expensive
Regarding rare earths, we will need them regardless of using nuclear or renewables for electric cars, so that seems partly tangential. For wind turbines their reliance on rare earths seems troublesome. It's geopolitically ironic that we (Western Europeans) get our gas from Russia and our rare earths from China.
You are right that you can't have an electric wire across the Pacific to get power from the US to the EU, but more because the EU doesn't border the Pacific. On a more serious note, Novia Scotia to Ireland is about 2600 miles, so it's on the edge of doable with HVDC, but it would be a massive project and slightly hard to justify.
I'm interested to know what rare earth metals make up so much of a wind turbine. According to https://www.nrel.gov/docs/fy17osti/66861.pdf (page 65), around 99% of a turbine is accounted for by steel, fiberglass, iron, copper, and aluminum.
https://www.offshorewind.biz/2021/05/24/usd-30-billion-north...
Whats funny about this is, that the wind farm will not even be connected to Iceland itself! I guess they have such great geothermal resources that they don't really need more power.