In the end a mix of all of the below:

- overprovisioning because it's still cheap enough

- peak shaving with battery storage

- dynamic demand applications, powered by hours of "free" energy, taking them outside the peak demand: car charging, heat and cold buffers (aluminum smeltering, ice buffers for A/C), and lots more that will prop up

Also

- mix of sources. Solar+wind is more reliable than either solar or wind.

- geothermal

- interconnected grids. The wind is always blowing somewhere.

Your third item implies massive multi-trillion dollar upgrades to the transmission system(s) which are historically sexy and politically unpopular. They need to happen but they won't.

It is a long bet on renewables+storage. From easiest to hardest (and most efficient to least efficient):

- Renewables+batteries are already cheap enough for a few hours storage

- Flow batteries can go overnight to weeks

- Synthetic fuels (i.e. hydrogen, or synthetic hydrocarbons made with green hydrogen & CO2 captured from the air) can be stored indefinitely and transported globally

Those are all possible, but according to the DoE 95% of large scale storage in the US is pumped storage hydro:

https://www.energy.gov/eere/water/pumped-storage-hydropower

It is also popular in Europe.

We can actually store energy from the grid. Not as inexpensive as generating it, but not unreasonable either. Most new solar/wind installs are now coming online with approximately 4 hours of battery backup.

Currently the shortfall is dispatched with conventional sources (mostly gas, coal and nuclear similar amounts). Battery prices continue to fall. Also new large storage facilities are coming online, such as compressed air which promise tens of Gigawatt-hours of storage at prices below battery. So that's the long term plan.

Actually if everyone had a personal power storage system they could probably reduce their bills alone with just that. Buy the power during the cheap hours and then use it later in the evening when it's 10 cents per kw more.

The end game is probably going to be a mixture of storage and moving energy intensive processes (e.g. aluminium smelting, hydrogen production, car charging) to when electricity is most abundant.

Smelting doesn’t work that way.

Aluminum smelters do everything in their power to avoid a situation where they lose power for more than a few hours and their pots freeze.

So you can dispatch for a few hours, but it adds enormous operational complexity if you can’t depend on 24x7 power.

Depends on where in the world you are looking. For example, the EU has a price on CO2 which is going up to reach the EU goals with respect to CO2 emissions.

In that case, even if storage of electricity is expensive. At some it will be cheaper than burning fossil. Nobody really knows how new nuclear competes with storage.

If there is no price on CO2 (and now other regulations that limit fossil), then renewables just have to compete with fossil on price.