The Ogre TinkerPop3 integration is still under development in the master branch. TinkerPop3 creates less clear lines as it pertains to where components begin and end. In TinkerPop2, we had this clear separation between Blueprints, Gremlin, etc. Now it's all "just Gremlin", which blurs the lines for libraries like Ogre which used to just be about graph traversal (i.e. the clojure flavor of gremlin). imho, a first release of Ogre on TP3 should likely just focus on providing clojuresque functions for dealing with the Gremlin APIs for "structure" and "process":
The GitHub issue tracker shows what's left to be done with Ogre for it to be caught up to the TinkerPop3 API which is still seeing some change in its milestone releases:
Give the incomplete state of Ogre at this time and the major changes TP3 introduces since TP2, I would recommend learning TP3 natively first. Once you have a reasonable feel for the TP3 API then take a look at the code for Ogre. It isn't too hard to follow - i think even those new to clojure would find it easy to jump into.