- the new shader arc needs to perform as good as the old WLIV in bruteforcing applications (so we can get the projected 2x performance based on the shader numeber)
- drivers don't cause issues, existing apps work on the new cards from the getgo
- prices aren't outrageous
- IF performance is indeed a 2x jump, they have to come out before the increase in difficulty invalidates their power from the getgo (so it is economical to invest in the cards - a change from, say, 0.02 to 0.04 btc per card per day won't make it worth the upgrade on current mining rigs, unless btc value jumps to $300 by the time the cards come out).
Also, the previous run to $30 was caused by some wacko buying over 3000 btc in the right moment, jacking the prices up by $4 and causing speculators to buy en masse lest they miss the gravy train. To cause a jump of $4 right now, one would need to buy waaaaaaay over 15k btc. So unless all sellers back out momentarily, I don't see how a run to $30 will happen soon.
But, mtgox only just reopened, so I have a feeling that a real rush has yet to happened. Fun things may happen by thurdsday/friday, once wire transfers go through to dwolla. Maybe.
Lots of people with free trades = high liquidity = stability and sanity in pricing