Do you think its reasonable to think that game benchmarks done on fresh installs do not need as many cores as a system thats running skype, antivirus and maybe a few other programs in the backgroud. ie the i3 2100 wins most benchmarks against a Phenom x 4 955BE but if there are backgroud processes running then would the 955 be better and by how much? Another example would be does this situation make there a reason to get a 2600K over a 2500K in a pure gaming rig?
In my own experiences none of my systems have had quite the results listed in benchmarks for the very same hardware/software. While some have been very close none have been as good. I assume it is due to all the added background programs I have running at any given time. However even with everything running I am only using 1-3% CPU when idle and less then 10% memory. So it would seem that it can make a difference but enough of a difference to require more cores/hyperthreading/upgrades with todays current hardware? I just cant see the need unless you just gotta have that extra couple fps. I think the result would be minimal.
If you're running some antivirus programs (especially older Nortons & McAffees), it can significantly slow down your PC where the benchmarks really would get impacted. But even then, I doubt many situations (other than running winzip while gaming) will result in the 955BE catching the i3-2100 without overclocking it.