My real trouble is the probability of a high and safely sustainable overclock which is seemingly going to be hit and miss worse than ivy. What all these posts are not telling us is the increased probability of getting chips that will not OC as well as the engineering samples shown which is exactly what many have experienced with ivy in the near past. So I'd rather the ivy devil I know in this case. If the gains are 5-7% ivy to haswell what makes you think that OC clock for clock will yield anything better if not worse due to other factors that don't scale linearly in the real world? It still isn't worth it as the gains in OC are too measly to even make any difference in titles that matter like BF3 and Crysis 3 and warrant the premium that they are going to be selling at when they finally touch base in a store nearby you.
There is always going to be a trade off with an intel chip from the times of Core 2 architecture to present. If it is not the heat limit it is the OC potential if not that it must be the vcore required if not that it will jump back at the heat being enough to degrade your chip rapidly or you just have more trouble with the lottery than previous architecture.