At first, this may be very true. But 6 months down the line, no. Growth would only resemble this until all the initial preorders for all manufacturers are filled. After that, it depends on how many new users bitcoin attracts, the value of a coin, and whether or not "average joe" is willing to purchasing mining equipment.
I'm not saying we won't get to a 1PH/s on the network, but anyone who says it'll happen by the end of this year hasn't really done any statistical research into Bitcoins mining base. It's sort of like the switch from CPU to GPU...it's a massive gain at first, but ultimately plateaus in comparison. I imagine the switch from GPU/FPGA to ASIC will happen about the same way.
Where did he say he was going to maintain 27% of the network at all times?
1. Agreed. I anticipate a certain amount of overbuying or ASIC saturation in the market, though. To what degree I'm not sure. Such a condition will certainly kill network growth.
(The numbers in the OP are actually determined using an algorithm that follows sales volumes and ROI terms - of course, my algorithm could be optimistic)
2. He didn't. This is entirely for speculative purposes. Specifically, in attempt to understand exactly (if even possible) how aggressive ASICMiner will need to be in the future in order to make
.35 - .4 BTC / share a worthwhile deal.
Actually, it's very concerning to me that ASICMiner does not plan to be aggressive in self-mining. A shift in business model to consumer sales opens the question: how well will ASICMiner perform in getting market share away from the likes of BFL.
And then, a consumer sales based model, will introduce cyclical revenue patterns related to consumer ROI.
But ultimately, when per unit ASIC prices drop to near mfg cost, the better business model is to sell to a consumer rather than to host and operate an ASIC farm.
So, my concern may not be too great of an issue. We would just hope for good decisions on ASICMiner's part.