As long as the cost of the attack is affordable by states, states can launch the attack. Always remember cryptocurrency is much more an economic issue than a computer science issue and the overall attack vectors are far more than the vectors from computer science; a computer science expert says that Joe's email password is strong enough to break in 1000 years but then Joe always fails in the attack in a social way by phishing or bribe.

Hashing Power Attack

One needs to purchase sufficient mining rigs to launch this attack. The cost of this attack can be estimated by:

1. H the overall network hashing rate
2. Q the price of mining rig
3. h the average hashing rate of mining rig

The cost is

Nodes Seize Attack

One needs to conduct inspection for possible nodes among the internet and seize them. The cost of this attack can be estimated by:

1. m the average internet speed
2. L number of computers in the world
3. S detective agent salary per year
4. w investigation weeks of a potential node
5. M the required internet speed for a node which is related to First Law of Cryptocurrency

Assume exponential distribution for computer's internet speed. This is the maximum entropy distribution with knowledge of positiveness and average mean. With the weakest assumption, it is more conservative compared with a Pareto distribution assumption; see here for the difference between the exponential distribution and typical Pareto distributions.

When states have no idea where the nodes are, the number of potential nodes in eyes of states detectives is so the total cost of the attack is

This also settles the block size dispute for solid number because of its relation with a proper internet speed. The logic goes like this:

1. The attack cost shall be large enough so that states don't want to do that given the limited budget. Favorably it shall not be less or greater than the cost of Hashing Power Attack because states always choose the cheapest attack vector.
2. By the reasoning , we conclude the required M.
3. We allow only 0.01 of energy waste in mining and orphan blocks , therefore we get the e. Alternatively, we can decide the optimal e by OptimalBlockDelayTimeToBlockTimeRatio.
4. By , we get the B
5. By , we get the N
6. By , we get the f

We can do no more better because of the First Law of Cryptocurrency. With this B, only transactions with fee greater than f is feasible. We feel blessed if our preferred transactions are covered. We feel compromised if our preferred transactions are not covered; don't curse anyone if we can not use this cryptocurrency for coffee.