What a crappy test. They've got all the resources, yet they perform just a few tests. Makes it seem almost like they were paid to make Fusion win...
agreed, iirc, in Fusion, it defaults to single core, and you can tell it how many cores to use... so why didn't they simple test Fusion twice? single core and multicore.
like fleurya says though, doesn't matter how many more tests they run, Fusion will win purely because it is a single core vs multi core test. and for that reason alone, that test and many others are just information on performance on single and multi core, until they test Fusion using a single core or Parallels release a multi core version, the tests mean squat as a comparison between the two.
what people must also remember, by giving Fusion access to both cores, OSX could potentially suffer far greater than if you were running Parallels or limiting Fusion to a single core, obviously depends on what you are doing on both OS's.
at the end of the day, each to their own
the performance of Parallels is more than adequate for me, and aesthetically I prefer Parallels over Fusion, and Smart Select is also a clincher for me
forget what the tests say, just use what you are happy with