Originally Posted by fivefeet8
They are both post process effects, but beyond that no. MLAA works by taking high contrast areas and blending them together for a best guess approach to finding edges. SRAA requires subpixel information so it needs access to some type of depth buffer to get that information.
This also means that it should be superior to MLAA especially in case where polygon edge is vertical or horizontal, in which case MLAA doesn't do a thing.
Most likely it should handle small polygons better as well, although it will be interesting to see a case of very thin lines where only every 5th pixel are actually shaded and the depth buffer tells clearly that the surface is there..