Welcome to the NFL (a phrase I've used before)! This is a situation not at all uncommon in the commercial startup world. The promoter (Nicholas Negroponte) has, by dint of impressive skills of persuasion, assembled a small but driven team that has created something sufficiently like the promised product to push out the door.
The investors, who previously stood back and watched, have realized that they may in fact see some return for their money, so they send in the experts they've used before to bring order to the situation.
The promoter loses his position as CEO ("kicked upstairs" to a lofty-sounding position with no power) and the ranks of management are purged and replenished with people who are generally more motivated by social advancement within the office than by the opportunity to innovate.
The company targets an easily-identifiable market and the product design is re-focused on that market. With luck, it becomes a cash cow, at least until competition catches up. The intelligent thing for the innovators to do at this time is to re-group outside the company and work on enhancements, applications and systems centered around the original product concept.
Obviously, adaptations will have to be made to ensure that these secondary products can be used with the base product as it emerges. They shouldn't try to ride the beast as it grows from a pony to a dinosaur (take it from one who's been there). To do so is to invite being corralled and separated from the possibility of creating something really new.
Steve Jobs with Next and the Donna Dubinsky/Jeff Hawkins team with Treo are successful examples of this step-outside-and-keep-innovating strategy, which they used to get back into the good graces of their original company. To put it another way, when I interview engineers I always ask "how would you have done it if you had been allowed to do it the right way?"
So Mary Lou's Pixel Qi approach is the right way. There is a lot of opportunity in the software and courseware field, too, as well as security (Ivan?).
Now making the sale (of the product or of its concept) is no longer a missionary sale, but can be done on the basis of comparison. The nonprofit structure of OLPC provides some interesting opportunities as well. If it survives, it will have enough of a cash flow to reward its management. Nonprofits can collect money for licenses without tax consequences and can dole out the not-profits as they see fit without the threat of stockholder lawsuits.
The structure is well suited for a cash cow product to serve as catalyst for more innovation. Remember, what remains to be built is not a company but an industry. In the personal computer industry the original innovators (whose names you will never hear outside the historical accounts) often became fixated on building their own cathedrals while the rest of us went around them to build the thriving bazaar (thanks to Eric Raymond for the metaphor). They were soon engulfed and their plans for domination of the industry quickly turned to dust.
Things are turning out better than I had feared. OLPC will become a structure with high inertia, turning out a product to be bought by slow-moving institutions while the founders and others generate a vital leading edge that is constantly growing and changing. The money with which to do the actual educational research can be spun off from OLPC to be applied by externally-assembled ad hoc teams working at high efficiency. Money will be attracted to fund new companies to carry improvements into the market and enable its growth.
A lot of us reading these pages will be the ones to exploit their corner of the opportunity, and in doing so, generate more opportunity. The crucial element is communication - across enterprise and institutional boundaries. Things have never looked so good as they do now for the successful development of low-cost systems to empower people in developing areas.