When it comes to overclocking potential I'm always paranoid after reading the first few reviews from the big hardware sites... I'm convinced that at some point in the history of benchmarking, sites have been sent pre-tested parts that represent the very best the company has to offer... so while Tom's e6300 would reach 3.6 GHz or whatever... the ones we buy will have more humble overclocking abilities. I'm NOT saying that's the case with the C2D (trust me... I WANT ONE!) but I do believe it happens.
One site somewhere had some unusually good RAM on a test bench that was sent directly to them... they go and order the same product on their own and it has considerably less OC headroom. Coincidence? You tell me.
When it comes to overclocking potential I'm always paranoid after reading the first few reviews from the big hardware sites...
Well, I've heard this in the context of Intel CPUs so I assume it probably goes for other parts too, but they have what are called Engineering Samples (ES) which are probably what are given to the media for pre-release benches/testing. These ES models are generally made in small patches and are much more quality ensured than your standard fare, thus you will see such things as increase overclockability.