Significantly higher temperatures aren’t what we’re talking about. We’re talking about a 10-15F spike.
Because heat flows at a rate proportional to the delta, it’s useful to think in terms of the delta between outdoor temperatures and desired indoor (let’s say 70 to make the math easy). So Phoenix has a few 120Fcdays a year, so delta of 50, and we’re talking 135F (so delta 65) for a 30% increase.
Most peoples AC is not sized to handle that, so it already runs basically full stop on the hottest days and can’t quite keep up. Maybe your home is 75 instead of 70. Making the exterior hotter has no effect of your power consumption once your AC is running full blast, it just makes the interior warm, so some people are going to have 80+ degree houses. They’ll be annoyed but won’t die and the power might spike let’s say 25%.
The power grids in these places are equipped for that because it’s life or death. You never hear about Vegas or Phoenix having a blackout on a particularly hot day because it’s not unexpected. Same with cold days in Minnesota. (You hear about it in California but that’s due to mismanagement and often unrelated to weather anyway.)
Now in Dallas when they froze you were looking 0F temps or thereabouts in a place where average winter temps are 56 and peak lows around 40. Further complicating it was that supplies of fossil fuels for heating ran out because it went on for so long. So you went from let’s say a 30 degree delta to 70, and electrical demand even higher than you’d expect when they ran out of natural gas and started plugging in space heaters. So you’re looking at like 2-3x previous peak demand, rather than 1.3. Nobody’s grid is prepared for that.
So again I’m fairly confident that if Phoenix hits a peak of 135 it is going to suck and a few hundred people will die for sure. It won’t be millions.
Global warming thankfully grows on 1 degree per decade kind of time scales which sucks ecologically but isn’t faster than we can grow the grid.