One could look at this as a one transaction per block blockchain with no difficulty adjustments. If this algorithm were used in a highly decentralized environment it would be extremely difficult for separate clients to reach consensus about a transaction history.
What would end up happening, because it would render the most consistent transaction history, is transactions would be sent to a node or maybe a handful of centralized nodes that would process the transaction.
You reference this problem here, emphasis mine;
One of the problems with the DAG approach is how to limit the maximum cut of the generated DAG
or, in other words, how to prevent all new transactions from referencing the same set of parent
transactions, and degenerating the DAG into a star graph. The DAG must not increase in “width”,
and it must “look” more like a yarn under microscope. I will call this structure a DAG-chain.
And then you say you address it with the following incentive structure;
DagCoin tries to fulfill that premise, using an incentive structure such that:
- There is a benefit for users to reference as many previous transactions as possible
- Referencing many previous transactions is incentivized only when there are many previous
transactions unreferenced.
- There is no competition between users to reference a previous transaction.
The first incentive is what is creating the problem. A node wants to get as many transactions as it can, and those transactions should reference themselves as much as possible.
The second, requiring that many transactions be unreferenced, is putting the cart before the horse. Nodes can't agree on what transactions are unreferenced before already having consensus about the current state of the network, which is what you're trying to solve to begin with.
Even if it wasn't putting the cart before the horse, only the most centralized nodes would have the most information about unreferenced transactions.
The third is irrelevant. Nodes can't compete to reference the same transaction because that would require they have the same information, which again, is the problem we're trying to solve in the first place; Consensus about the data.
So I don't see how this protocol addresses the problem of "degenerating the DAG into a star graph" as you put it, or being subject to extreme centralization pressures, as I would put it.
https://bitslog.files.wordpress.com/2015/09/dagcoin-v41.pdf
