I agree that the idea isn't too well thought out, but I think the idea is more to fix the ratio of difficulty to reward. Thus, as the difficulty goes up, so does the reward. As difficulty goes down, reward goes down. Difficulty could still be adjusted to maintain a steady 10-minute block window.
Thus, when the price spikes as it recently did, it becomes incredibly lucrative to mine and sell bitcoins. Everyone and their grandma starts mining, driving difficulty and reward both way up. The increased reward causes bitcoins to flood the market, putting downward pressure on prices. A similar pressure happens in the opposite direction.
The problem of course is if you are trying to create a stable coin there are extended periods of time when the appropriate thing to do is to generate ZERO new coins. That makes mining unprofitable for everyone,so everyone stops. In a bitcoin style framework that means that block creation gets seriously delayed and transactions don't get confirmed. After a while difficulty drops to speed the process back up, but the supply&demand exchange value hasn't necessarily fallen with difficulty so coin creation starts again. Even though the correct monetary policy may be to still create ZERO new coins.
As far as I can tell this doesn't produce a stable currency. It osculates tending toward a zero coin value. (continuous over production of coins) It also tends to drive the difficulty level down along with the coin values.
Your analysis seems pretty sound. Before I start agreeing with you, let me propose one change (at the risk of going out to la-la-land):
Let C and D be some constants, their value TBD.
subsidy = C * difficulty
This is the original idea, keeping the ratio of subsidy to difficulty constant. Now also, to deal with the times when the subsidy ought to be zero (and is so low that it's effectively zero):
mandatory transaction fee = D * moneysupply / subsidy
(It could equivalently be D * moneysupply / difficulty, with a different D value)
Thus we would have:
1. During "boom" or expansionary times, there is much demand for the coin and the exchange rate begins to increase. As the rate increases, it becomes more profitable to mine, so more people do it, so the difficulty increases, so the money supply increases and eventually applies downward pressure on prices.
2. During slow times, the exchange rate begins to drop. Mining becomes unprofitable. Less people mine. Difficulty decreases, and so does the subsidy BUT, counteracting this effect, mandatory transaction fees begin to increase, to maintain a baseline incentive to keep mining.
This might mitigate the effectively-zero subsidy issue.
Now, switching back to agreeing with you; here's another problem with the design: when a new mining technology appears that allows people to generate a lot more coins with a lot less electricity (e.g. ASICs), there will be a sudden major inflation. If the ratio of coins generated to electricity consumed increases by a factor of 100, I see no reason the money supply wouldn't also increase a hundredfold. At this point, the miners have most of the coins (they just generated 99% of all coins generated to date), and transaction fees are now 100 times higher. Whoops. We were s'posed to be getting a StableCoin, but instead we got a ZimbabweCoin.
