
December 17, 2017, 07:49:57 AM 

From a purely mathematical point of view (no software required) There is arguably $200 Billion worth of CryptoCurrency, but $200 Trillion worth of money in the world. So, at the current rate of growth of bitcoin for example, and assuming all other crypto did not exist, and bitcoin values are not really a bubble, and that there is a factor of 1000 (Go from Billion to Trillion) for a growth rate that would make paper money obsolete, then finally it would take less than five years for bitcoin to be the only money left.
So the answer should be 1000 / (5 years) for maximum possible growth under above scenario , and if say there were five leading cryptos after 5 years , then it would be required to grow one fifth of that. Also factoring in that the current block size is more than enough for the current transaction rate , and you will actually require even less growth of the block size. There is also another variable that makes this whole analysis in error and that is something in economics called "the speed of money" , and I understand it simply as if you double the money supply you quadruple the speed of money , or in other words you quadruple the total number of transactions per unit of money. So we will have to bring out the calculus books, and ultimately it becomes an economics problem not really a software problem to know exactly when to do it and how large it should be. That may be why there has been alot of disagreement about the issue within the core dev community.
From a software point of view , I think, if we could know the exact future of crypto rather than be agnostic about it , then calculating the most efficient blocksize would be easy.
