The more the cost of BTC rises, we can expect the expenditure by miners to rise equivilently, which in turn makes an attack even more expensive. By this definition, there is no such thing as getting the same security for a cheaper cost, because it's the cost itself that defines the level of security.

Also, I disagree that the $650M is supposed to "purchase decentralisation". What it's really purchasing, if anything, is security. In the context of bitcoin, we can measure security by the amount it would cost an attacker to mount a 51% attack. Right now, that's ~1800 BTC per day, or nearly $1M/day.

 And since we will have so many millions, or billions of dollars worth worth of new coins mined in a year, the equivalent value will be invested to mine them until the cost of production becomes only marginally profitable. Whether this means spending money on large mining data centers, or funding millions of sybil nodes to influence a voting system, it will happen either way, because if there's money to be made mining, everyone will do it until it becomes unprofitable.

The inflation schedule is a non-negotiable. Are you suggesting that we change the 21 million coin cap, or change the halving-every-four-years schedule? Because if we're keeping those rules the same, then the value of new bitcoins created must be <number of coins> * <bitcoin price> per year. No change of consensus algorithm will change this.

"818k followers", but 131 retweets. The retweets is a much better metric of interest, and 131 retweets is absolutely fuck all.