The render and display plugins would be where you'll have to put 90% of the code, but you still have to add 3D data to camera (3D frustum/position/orientation), graphic (mesh/materials/skeleton) and anim (joint keys).
There wouldn't be much to write there as those modules are only storages for abstract data. It's the render plugin that actually knows what type of data is stored by graphic & anim and how to handle them and the display plugin how to display triangles and materials.
Comments
The world where objects live is in 3D.
Cameras, graphics and animations have been designed so that they could accept other types than 2D data (but that requires work to implement them).
The only available renderer plugin that has been written so far is a 2D one.
There wouldn't be much to write there as those modules are only storages for abstract data. It's the render plugin that actually knows what type of data is stored by graphic & anim and how to handle them and the display plugin how to display triangles and materials.