I forgot to just point out, that a simple fork, like golos is to steem, does enable the elimination of issues like voting prisoner's dilemmas, but I am more concerned about the mingling of the data types within the same graph database, there is a big problem:

https://neo4j.com/blog/dark-side-neo4j-worst-practices/

Another common mistake I see is using Neo4j as a Binary Large Object (BLOB) store. You might have heard that we do support byte type properties, and we do support arrays of primitives so byte arrays are possible. This, of course, opens up the possibility of storing gigabytes of data in a property value

Binary or not, large blobs of UTF-8 still has this same problem. Steem, keeping the forum data with the rest, in a graph database, leads eventually to an unsustainable load on the indexes and IMHO is already happening.

Never mind the fact that it is

• written in c++
• uses boost, which seems to change api more often than babies need nappies
• uses a graph structured database on simple linear tables that rarely use more than two tables, or would be, were they stored as tables.

The designers of blockchains are pretty much ignorant to database theory, and graph databases are for complex data with many lateral relationships.

This is about as far away as you can get from a double entry ledger. There is no sane reason to use graphs. It would be pretty neat for generating searches for full text of posts. But that isn't. Google is steem's outsourced full ext search. and it is useless...