So one difficulty is that the future price of electricity is hard to predict. There is a whole lot of money that goes into trying to make probabilistically accurate predictions about the future prices of things. If it were easy then there wouldn’t be such a large industry around solving these problems. Because it is hard to predict future prices, it is hard to know if prices have temporarily changed (ie if prices are high/low) or if prices are going to continue changing and will stay that way for a while (ie if prices are currently lower/higher than they soon will be).

For example, if you had a smart home that cut off heating in the recent Texas event when it saw prices rising (because it assumed prices would revert soon allowing for cheaper heating a little later) it would end up having to pay even more later when prices turned out to continue rising.

I don’t really think smart appliances are a sufficient solution for ordinary people because there is too much power usage that is fixed (or worse, correlated with high demand/low supply) and I don’t think consumers are actually signing up for the price volatility on their fixed usage.

It feels like you are suggesting that it’s ok to have everyone picking up pennies in front of steamrollers because they can speed up a bit if the steamroller gets too close or go and pick up more pennies if it is further away. But I say that this won’t be good enough if the steamroller starts going faster and they get too tired running away. It’s especially bad because many people merely thought they were picking up pennies and didn’t realise they were being chased by a steamroller. I feel like this analogy is pretty strained by now but the recent events in Texas showed a lot of people getting squashed without realising that they had been spending years running around in front of steamrollers.