It's most correct to plan around every location getting more wet, more dry, more hot, more cold, more storms and more droughts.
Remember that humans are actually very adaptable, our range of habitats extends from the arctic to the tropics, desert to rainforest. Most life is not so tolerant.
(I’m not actually concerned with us running out of cheap abundant energy; PV is already fantastic in this regard, and in a hypothetical collapse of all dispatchable power sources that included never being able to make more batteries, we would be severely inconvenienced rather than utterly finished).
This doesn't really track. We're talking about adding less than 1% increase in energy.
A large variation, yes. A direct relationship between a small increase in input and a corresponding large increase in a specific output, no.
Sure, which makes the blanket statement the GP made incorrect:
>Putting more energy into a meta-stable chaotic system will greatly increase the number of extremes across the board
In either case, a 1% change in energy input is always going to be a big deal with regards to an energy dependent chaotic system.
Chaotic systems, in essence, are known for its unpredictability. Watch this to get an idea: https://www.youtube.com/watch?v=pEjZd-AvPco