It's actually not that fancy. It's basically a fuzzy arbitrage with risks included. For a small investor, to be able to get to the "efficient frontier," you have to make a lot of assumptions none of which you can actually validate. So I don't believe anyone here is using it.
I don't doubt that some big investment firms that employ supercomputers might be able to do it but I doubt it. The evidence is the screw up by the LTCM. They certainly understood "efficient frontier" but boy how badly did they miscalculate!