Well, if a million people use the Hive blockchain then the price of Hive would be way up, so it would pay for much more powerful servers to be run by the witnesses.
Right, so what about every other node that's not in the top 20?
If it becomes prohibitively expensive to run a node we will only have 20 total.
This greatly centralizes the chain and leaves us open to multiple attack vectors.
It also kills a lot of projects that would have otherwise done well.
To be fair I think we are doing a pretty good job with the scaling situation.
But many people around here don't grasp the technical challenges of scaling up.
We've all simply been told that Hive scales better and most users blindly believe that.
We could do the math. IF the earnings on much higher hive price would be enough to cover the cost for top 100.
Also services maybe host a node for there game/app, that generates enough by itself, so they don't need to cover 100% on witness earnings.
I think only longterm would show up. Worst case, cut a bit top 20 and share it with top 100.