Some key insights from the article:
Basically, what they did was to look at how much batteries would be needed in a given area to provide constant power supply at least 97% of the time, and the calculate the costs of that solar+battery setup compared to coal and nuclear.
What you do is get weather data for sunlight and wind. The two combine to cover some of the lull in the other. From historical data, you can calculate the maximum lull where neither are providing enough. Double that as a safety factor, and that’s how much battery you need.
Doing this is by far the cheapest way to get to 95% clean energy everywhere. That would be a total game changer.