The price is determined by two factors:
1) The cost of manufacture.
2) The price the consumer is willing to pay.
re:1)
The cost of manufacture will not be that high. Current 45nm technology can produce 4 core 2.66 i5 cpu's that sell at $196 in industrial lots.
32nm technology can put that same cpu in half the amount of silicon. If you extrapolate that, a similar 6 core cpu should cost about $150 for the silicon manufacturing part. Assembly and marketing costs will be similar, so the cost to Intel should not be much different than the current $196.
re:2)
What would YOU be willing to pay for such a 6 core cpu? What would enough others pay? If it is less than $200, then Intel made a bad decision, and will not produce the cpu. Not a likely prospect. There will be a significant group who want a 6 core cpu @2.66, even without overclocking. If overclocking is as easy on the i7-920, then many more will be willing to spend on such a chip. Considering the success of the i7-920, many will be willing to pay the same $300.
Even now, the top of the line chip fetches $1000 and some are willing to spend that. Intel will price these cpu's at whatever they think will maximize their revenue. Over time, the price will reduce, once the high demand early adopters have paid the price. It will do well in the server arena where more cores are needed, and the price is not a big obstacle. The server chips will be labeled Xenon or somesuch. What is unknown is now much 6 cores are valued in the desktop and workstation market. At this point, gamers can mostly be satisfied with two fast cores, and certainly not more than 4 cores. Will they pay a premium for 6 cores? Some who demand the best will, but I think most won't.
So my conclusion is that we will see 6 core cpu's appear in the same price range as the i7 920-975. Because of the 32nm manufacturing, I expect the clock rates to improve some, perhaps 10-20%.