Only if your running multiple threads will HT help, for the HT to take affect the software has to be able to utilize it. There was some basic benchmarking I was a part of, nothing scientific but still pretty straight forward and the consensus of the results were that disabling HT increased performance slightly encoding a single file using TMPGenc.
Although completely unscientific what was good about our tests was just that, they were tested like a normal user would use them in the real world. One of the biggest surprises was the wide range of results even from the same CPU's or in my case being beaten by a slower CPU.

In this case all of my specs were higher including the RAM and we were both running mobo's with identical chipsets. The major difference was he was using a better mobo than myself which lead me to believe that my bottleneck lie in the mobo somewhere. Just felt I would point that because there's more to it than putting CPU to CPU in the real world.