AigoMorla -- This is just "intelligence" from second sources, but there are enough of them to increase the probability that the "skinny" is accurate.
A lot of customers posting reviews at Newegg since July 22 are saying they got the G0 stepping. On the "market-economics" side of the issue, Newegg was offering the C2Q Q6600 for $298 on July 22. Then, several days later, the price went up to $348. [I understand that MSRP was supposed to be closer to $250.] A few days after that, the price went down to $319. But it was during the interval between the price-rise to $348 and the decrease to $319 that these people reported getting G0 steppings.
I won't go into the behavioral logic followed by Newegg, except to say that -- if they're successful, if the continue to "survive" -- they follow Day's Law in the long run. This assumes they are in a near-perfectly competitive market, and that's a safe assumption for sure. They'll look at the number of fish that take hooks with a certain amount of bait on it, and weigh that against their unit profit margin, then adjust the price again. There's even a flow-chart that describes all of it. They'll also attempt to get rid of older stock at lower prices while minimizing losses, or capturing as much profit as possible before it's too late.
I'll take a look at the Tuniq in the comparison reviews, but I'm not very fond of Sunbeam-Tech right now. For performance ignoring certain inconveniences, the Ultra-120-Extreme is still tops, unless any of you can show me some new models and manufactures that have just been released with comparison performance benchmark reviews to back them up.
Someone asked about what you could expect stock temperatures for a Q6600 would be (I thought he meant the Q6600 and not the Q6700). With the B3 stepping, your temperatures are likely to be higher than these, although these are obtainable on air with the TR Ultra 120 Extreme**:
Q6600, stock settings with Vcore Fixed at 1.28V Intel retail box maximum = 1.35V
The quoted room-ambient is about as close as I could measure it with two analog thermometers. It could be off by +/- 1F degree, but no more than that. This would adjust the temperature distribution up or down at most a half-degree Celsius.
So far, people I have asked to report their peak load temperature values (upper bound of the range -- for my distribution, it equals 52C) -- were operating in room-ambients between 73F and 75F, using identical speed settings and VCORE values that didn't differ from mine more than about 0.02V. Their peak load value or range upper-bound was closer to 63C to 65C. This value, of course, is the peak load value shown by the hottest core measured in CoreTemp or a similar program, and there is no guarantee that the programs report identical temperatures, although it is unlikely there is significant deviation.
** Tuniq might come close, since these are stock settings, but I'll wager $10 the idle-to-load spread will be greater. I'll make a side-bet of $5 that there will be no significant differences between idle values. This assumes that the same fan and thermal paste are used in the comparison, but check the May 4 review comparing the Ultra 120 Extreme. I believe the Tuniq was included.