Reading the comments here, the results seem to fluctuate heavily in dimensions, winners and margins.

I guess, this test is just not an epitome of anything at all.