This has me interested as well
I can see the problems of hashing incentives leading to a centralized mint if the network, lower fee rates and higher transactions through mining more blocks leads people to favor some pools over others.
Such as Ghash.io
In relation to the blocks solved by miners higher discovery rates tend to equalize payouts, since I am not a dev can't really speak much on the solution other than somehow splitting the miners up by having a dual method verification system.
Such as Sha 256 and another algorithm that would work together on the ledger requiring different software to operate making a 50% attack twice as difficult, or a sidechain that does the same thing based on the Bitcoin protocol and incentivizes miners to use it as a way to earn revenue.
Simply put two reward pools one made easier than the other but in a way that promotes distribution and decentralization over centralized networks.
__
From Reddit
1) Eliminate pools.
2) Provide a way for miners to solo-mine with low variance and frequent mining payouts even with only small amounts of hashing power.
3) Get rid of ASICs.
__
Possibly
1) Split pool distribution in a way that incentivizes miners to not collect hashing power into one pool
2) Provide a way for miners to solo-mine with low variance and frequent mining payouts even with only small amounts of hashing power.
Perhaps with a sidechain implementation or a fork.
3) Removing Asics
Difficult to implement other than to change the scripting algorithm, an increased hash rate also strengthens the security of the network but other than changing distribution or scripts this is difficult to counteract. (That said more nodes are needed)
If a split and fork is supported would also implement 4 into the suggestions
__
4) Increase nodes and promote an incentive to running full nodes that keep a record of transactions occurring over the network.
As data increases and the blockchain grows running full nodes becomes more burdensome so a mechanism of reward however minor may increase the decentralized aspects redundancy is important.
http://www.coindesk.com/bitcoin-nodes-need/Anyways just my thoughts on this.
That said: A 51% attack should not be as profitable as earning the revenues from Blocks so the costs of doing this outweigh the benefits of such an attack.
However A centralized mint defeats the point of decentralization so it should be addressed.