The author's goal for this design seems to be to increase mining decentralization, but there's another big advantage to a blockless graphical ledger over a bloc

The author's goal for this design seems to be to increase mining decentralization, but there's another big advantage to a blockless graphical ledger over a blockchain. By attaching proof-of-work to each transaction rather than to a block of multiple transactions, a blockless scheme asks miners to answer a simple cost-benefit question: does the payoff to putting the transaction in the ledger (in transaction fees and seignorage) exceed its cost (in hashing resources)? Particularly when seignorage is zero, so that a miner's only compensation for providing proof-of-work to a transaction is the fee it offers, this eliminates a free-rider problem created by blocks in a scheme like Bitcoin's; as long as you can include as many transactions in a block as you like, the cost of including another one is nil, and so miners have an incentive to include every transaction that pays a non-zero fee, no matter how small [0]. That means the equilibrium price of putting a transaction into Bitcoin's blockchain, for example, may never rise much above zero. That's a real problem for the long-term security of the Bitcoin blockchain. Per-transaction proof-of-work eliminates that problem by making both the cost and reward of adding a transaction to the ledger a function of the properties of each transaction, rather than of an arbitrary aggregate of transactions. 0. This isn't true if there is a binding limit on the size of blocks, as there is now in the Bitcoin protocol. But a wholly arbitrary block-size limit is a terrible solution to this problem because it is unresponsive to economic factors. It may be possible that a non-zero equilibrium price of transacting could arise in response to other constraints (like the risk of "orphaning" a large block due to the increased time it takes to propagate around the network), as some people have argued in the course of Bitcoin's ongoing block-size war.