"knowledge graph Optimization" is a great term to use to differentiate between "rich snippets" or enhanced displays that directly extract structured information from a webpage itself..
Both are enhanced displays in SERPS, but where is the line drawn between Searching over data (aka internally stored and curated results such as the knowledge graph) and searching over documents such as cached, pre-indexed web pages. Are we able to differentiate it in SERPS? For now, I would say yes.
We now have results that "collide" with the knowledge graph, (i.e graphs of information that intersect /overlap or "collide with one another from an information point of view" & for which we need KGO) vs results that extend the knowledge graph from some hopefully/presumably trusted web source?
Agree with AJ about this being potentially huge and go with +Aaron Bradley
on officially coining the acronym.
If searching over data one day supersedes search over documents, where does the knowledge graph end and the web begin?