Proof of work only reaches Nakamoto consensus if the average value transacted per block is less than twice the amount paid to miners. Otherwise it potentially

Proof of work only reaches Nakamoto consensus if the average value transacted per block is less than twice the amount paid to miners. Otherwise it potentially becomes viable to 51% attack - it becomes cheaper to buy miner time to reverse my purchase of internet heroin than to buy the heroin. Thus, there is a limit to the amount of economic activity that can be "safely" transacted by a proof of work chain per unit time, based on the amount of inflation/fees of that particular network. The more fees and inflation, the higher the economic value that can be transacted safely. The actual real-world limit will be somewhat less than 2x because of options trading and other off-blockchain transactions. Yes, this poses problems for smaller currencies where the average amount transacted per block is quite small. Bitcoin has not just a "network effect" of adoption resulting in places to spend it, but also a network security effect, where the vast amount of spending makes it much more economically difficult to justify a 51% attack. (That or you need to wait increasingly long periods on these shitcoin chains before accepting them as "final" and transferring goods. Eventually all chains should reach consensus, as long as you wait long enough that it would be economically unviable to And note that this is actually a parameter that cryptocurrencies can set themselves. When (eg) Ethereum cuts the block rewards, that results in network hashpower going down, and thus reduces the economic throughput that is possible before nakamoto consensus falls apart. Same for Bitcoin's block halvening.