I have read this thread and continue to be amazed at the poor job our government schools do in teaching that there is more to how a computer market works than a rudimentary supply and demand diagram.
For example, there is the concept of "loss leader." The argument goes that Apple should sell below cost, or at least below this magic supply and demand intersection people speak of, in order to grow their user base. Lose some money now, make that money back later when people upgrade their Macs and Mac software. This loss-leader strategy also has the added benefit potential of increasing the number of developers for Mac OS X which in turn will make Macs more attractive to different market segments.
The counter-argument to this loss-leader strategy is that the demand for Macs is not very price-sensitive, meaning even if they lowered the price quite a bit, there won't be much increase in sales. Now this is something ONLY Apple knows because they don't break down the sales figures by different models. If you had individual sales figures, you could track how changes in price of a given model affected its sales.
Personally, I think Apple can gamble a bit with all that money in the bank. I'm not a corporate strategist so I don't know the benefits of keeping so much cash on hand but Apple certainly is not saving that money for a major acquisition. Apple is not the kind of company to, let's say, go out and buy cable company or something. That's Google. Apple makes small technology-oriented acquisitions and partners with big companies. So what's the point of saving all that money? Surely Apple can afford to blow off a billion or two (say through a $200 rebate across the board) to get some momentum to increase the market share.