This. Preorder = risk. Risk is the potential for profit. I thing people are going to be terribly disapointed once companies have units in stock for immediate delivery. Look at ASICMiner as an example. I mean once built a company can do the same math a consumer can. Future cashflow can be reduced to net present value. If a company has a miner and it has a net present value of 50 BTC why would they sell it for less than 50 BTC? They could simply use it to mine and generate >50 BTC. Now if a company runs out of space, or power they may sell excess capacity for slightly less than net present value to increase demand but they days of massive "potential" profits are over. 100%, 200%, 300% gains per year. Never going to happen with in stock miners. Maybe 10% or 20% if you buy at the right time, have the right climate (cooling costs), and the low electrical costs.
Once again look at ASICMiner blade prices. This is the same equipment used by ASICMiner. People complain the prices are so high that all the potential profit is gone. As the CEO of ASICMiner why would you see it for less? If they (using their own projections on difficulty growth) project a blade will earn X BTC (discounted for risk & time value of money) it makes no sense to sell it for significantly less than X. Shareholders would simply be subsidizing the profits of the buyers. As a CEO looking to maximize the return for shareholder it doesn't really matter where the revenue comes from (sales or mining) the goal is to maximize revenue.
That being said some companies (BFL) have taken this to an asinine extreme so some backlack is inevitable but like you I really don't think most people appreciate what "in stock" prices are going to mean. Another way to look at it is imagine you have a rig (or rigs) which produce 2 TH/s. What would YOU sell it for?