If there was severe deflation, as you describe, the reward would increase by 1 for each block. Assuming that there is a new block every few minutes, the reward could get huge very quickly. So I think the supply could respond quickly to sudden deflation. I don't know if it is necessary, but the increase or decrease in reward could also be expressed as a percentage of the coins already in existence.
If there is severe inflation, it would take longer for equilibrium to be reached, because the rewards would go to zero, and then the demurrage would gradually reduce the coin supply.
In both cases, the expectation that the price would eventually return to the target value would probably have a stabilizing effect. If there's an increase in TargetCoin price, it doesn't make sense to buy a bunch of TargetCoins in hope that the price rises further.
If there is severe inflation, it would take longer for equilibrium to be reached, because the rewards would go to zero, and then the demurrage would gradually reduce the coin supply.
In both cases, the expectation that the price would eventually return to the target value would probably have a stabilizing effect. If there's an increase in TargetCoin price, it doesn't make sense to buy a bunch of TargetCoins in hope that the price rises further.
That would mean miners would have to mine with very much changing difficulty. I wonder if they could exploit that. It doesn't make sense to mine on a huge diff, so i would turn off my miners then, saving power cost.
Then... why should the bitcoin price react that way to the difficulty? No normal person besides miners would think that it would be worth more then. Bitcoin price isn't created that way.
I'm not sure what you mean. Are you saying that if the reward increases, more miners will try to solve the block, which means blocks will get solved faster, which means the difficulty would have to be adjusted? That seems true, but it also doesn't seem like a problem.