2) the plan for the last 20% of renewables is a bit more murky than the first 80% which most seem to agree will be straightforward and inexpensive and using the obvious solar and wind tech. I'm fairly confident that a range of solutions will present themselves, I just wouldn't want to predict exactly what mix they'll be as it'll come down to a lot of small innovations and decisions. We have answers now but I believe we'll have more, better, cheaper answers by the time we need them.
3) Rare earths aren't rare, that's not what the name means. So they're not a fundamental blocker, though they are more expensive to extract so the market for electrical storage will naturally gravitate towards more cheap and abundant materials if they're out there.
Batteries for short day-to-day storage are already entering the market and making/saving money. As renewables expand and battery prices fall they'll expand massively (see my answer about EVs). I think we'll need something with a different model for long term seasonal storage though, but the main thing will just be building lots of cheap renewables. The "problem" will be as much about all the cheap abundant energy we have in spring/autumn as about the shortages caused by winter heating, summer cooling peaks. Storing that electrical energy as ammonia/hydrogen/whatever to shift energy is going to be a big thing I think, but only so far as it's cheaper than just adding yet another solar panel. (The current big innovation is just putting more solar panels on the back of your solar panels. It only add 1% but it's cheap, so why not?)
1) if we were to eliminate ICEs completely in the US, how much would more electricity would we need to produce?
2) Is there any declining returns of renewables as we aim for that much more? (I don't think but just checking.)
3) Assuming no problem generating the energy, how much more batteries would we need? How much more rare earths and other environmentally costly resources would we need to mind, or are you proposing e.g. fluid pumping?