This hypothesis makes a lot of sense. One should be able to test it on a Nehalem architecture: benchmark the program running on two cores, once choosing them to

This hypothesis makes a lot of sense. One should be able to test it on a Nehalem architecture: benchmark the program running on two cores, once choosing them to be different physical cores versus the same physical core. Your theory predicts the latter option to be much faster, and faster than the single core mode, because the two virtual cores will share the same caches.