When you present 2D or 3D as a visual stimulus, the affected brain regions are different.  

how you are going to use 3D flicker as stimuli and why?

I would guess that for SSVEP, shapes/meanings etc. of stimuli do not play an important role. SSVEPs appear as pre-cognitive evoked potentials, and are related to automatic retina response to rapid visual stimulation. You may want to verify, however, if e.g. changing from 2D to 3D can lead to an amplitude increase, or a similar effect.

I agree with Lukas, there sholud be no difference between using for example a virtual (2D) or using a real LED (3D). Something that could improve the performance of the 3D stimuli is the realistic environment, the subject could be excited because the experiment and its entusiasm should improve your results.