Nice example Danny. There has to be some investment cost offset which makes things very slightly less advantageous to those who adopt the new technology. Additional power costs as well unless the new tech is much more power efficient. Never mind the lost income from having money tied up in a pre-order for so long, or perhaps lost altogether (keep the faith BFL'ers!).
It is my understanding that ASIC are significantly more power efficient. But yes, there are capital costs in purchasing the new tech when you already have perfectly working old tech. As you mention there is also lost income from having spent money in a pre-order when that money could have been used to increase revenue during that time. My example used some extremes for the sake of making the effect obvious (only 100 miners, a single miner acquiring 33% of the entire network's hashing power, etc). In reality a person needs to take into consideration a lot of variables:
- What is the current total network hash rate?
- What is the future total network hash rate likely to be?
- What is the purchase cost of the device you are considering pre-ordering?
- What is the expected hash rate of the device you are considering pre-ordering?
- How much hash power could you immediately supply with current technology for that same price?
- How much electricity per hash is used by the current tech?
- How much electricity per hash is used by the new tech?
- How much do you pay for electricity?
- What is the current bitcoin exchange rate?
- What are your predictions for future bitcoin exchange rates?
- What is the risk that the future tech won't be delivered on time?
- What is the risk that the future tech won't be delivered at all?
- How reliable is either tech, and what sort of warranty is available?
- What is the re-sale value of either tech likely to be in the future?
I'm sure I'm missing a few, but clearly it isn't a simple answer and the answer will be different for everyone depending on their own predictions, expectations, risk tolerance, and costs.