Currently, the Geometry2DDataVtkMapper3D (the mapper for the 3D RW) gets the slice from the mitkImageVtkMapper2D and applies properties like color, opacity etc..
With the new 2D rendering this is no longer necessary. The 3D mapper can just get the complete actor from the 2D mapper and render the same object in the 3D position. Thus, a lot of effort can be saved and a lot of features which I implemented/fixed for the 2D rendering should also work inside the 3D RW. For instance, the RBGA levelWindow or segmentation contours should work then.