It's hard to say how heavy the impact will be. I don't know what percentage of the network hashrate is in China...but you can bet that a lot of those GPU farms are going to be switching over ASAP. So we'll lose some GPU hashrate, but we'll gain MORE in ASIC hashrate.
There's only so much hardware they can reasonably produce in a given timeframe, though...so I'd expect a big initial "pop" in difficulty and then some slowing over time, until the other hardware makers get theirs into circulation. Once the other major players are in the game, it'll be the Bitcoin story all over again.
The question is, does the difficulty go up high enough to make the KNCMiner solution a less profitable use of $10,000 when it launches? It's a substantially higher amount of hashrate, but it's really hard to say what the difficulty will be when it releases. They've given themselves until the end of September by saying "Q2/Q3" so who knows? Alpha Technology still insists they're on target for July release, so that's a couple of months worth of difficulty increases...but they still need to address the competition with their plans regarding hashrate.
You are absolutely right about Scrypt to be like SHA256.
I'm just looking at it this way. If history repeat itself then we can say that KNC sha256 miners released in Oct/Nov made ROI and earned some coins for their owners. Ant's from Bitmain were not that good as KNC was.
Now we have this 64mh/s or 75mh/s miner coming out and it might be the same scenario. Who ever gets in first will get something out of it by the time KNC starts shipping their miners.
My 2 issues are:
1,This is going to be first batch of these miners so there are going to be issues (100% confident about this)
2, Pre-order (Get in stock and then start selling)