I firmly believe that we won't see any consistent spectacular overclocks until Intel starts releasing 3.5+ ghz CPU's, and that stepping migrates down.
I consider this similar to the cB0 vs cC0 stepping for the 600e/700e chips. The first cA2 stepping of the 600e could only reach 756 mhz if you were unlucky, and 800-820 mhz if you were lucky. When the cB0's first appeared, there were tales of the 600e's hitting 900 out of the box, either with or without a minor voltage jump, and I bought one of the first cB0 600e's and mine hit 900 @ 1.7v. Some 700e's were reaching close to 1 ghz with very good cooling.
The cC0's didn't really improve very much, although cC0 700e's were reaching 1 ghz more reliably at default, while *many* 750 mhz cC0's were failing to reach past 900 without extreme cooling (!) which led crecedence to the theory that Intel was REMARKING 750's that passed the test, as 1 ghz's. cD0' 1000E's were a big boost, as some were reaching 1300 mhz with them; no idea about the slower chips, as the FSB would have to be VERY high...
The C1 stepping P4's should get better with time, and if not, perhaps the next stepping will.
And there is a BIG difference between getting 600E cB0 (many actaully felt these were remarked 800 EB's since virtually all of them reached 800 at default without a sweat), and 1.8 C1's that are overclocking to 3.5 ghz, which have pretty much been proved to be remarked 3.06's, versus ones that aren't.
Sorry if I sound confusing....