Fair enough, I wasn't aware performance was that bad.
Regardless, my point is that the article didn't specify if he meant logical or physical cores, therefore saying that the article is wrong, is partially incorrect (and seems like an unnecessary attempt to nitpick for the sake of nitpicking).
On average, a setup with 4 logical cores and 4 physical cores performs 15%-30% faster than a setup with just 4 physical cores.