As someone with some experience in:
1.) Marketing, especially hi-tech gadgets,
2.) Turning hi-tech gadgets into useful, usable products, and
3.) The education market, I'd like to make a few comments.
1.) Marketing: everybody lies. Many folks who say they want one won't buy one, but, if the product offers a good price/performance ratio, some people who say they'll buy one will actually buy 100. Or 1,000. Or 1,000,000. Expect your best estimate of market demand to be off by 1 to 3 orders of magnitude. In which direction, I can't say.
2.) If the Pi delivers what it promises, folks like me will snap them up to turn them into little embedded servers to do all kinds of things. I'd buy 5 right now (and will as soon as they're available); if they work as I hope, I'll start buying them 100 at a time.
3.) As someone commented, selling to "the education market" is a long sell - you can sell millions of units, but it will take many months, stretching into years to get any momentum. And you'll have to have demand "pulling" from the teaching community, which takes time and energy to create.
I'm going to get some and just give them to teachers and students I know to stimulate demand here in Sili Valley.
"What do I do with it?" they'll ask
"It's a $25 computer with USB and HDMI output, use it to prop up a table leg for all I care," I'll respond. (actually, I'll probably get the Ethernet model)
Of course, they won't... Most of them won't do anything interesting with it, but, if we get it into enough hands, some will. As it starts to drive papers and presentations at academic conferences, etc., demand will build for ones and twos, then tens and twenties, then thousands.
The key, of course, is getting your manufacturing up to volume (in a somewhat orderly way) so you can drive down the cost and keep "the machine" going on the cash it generates (this is true whether you're trying to create cash flow for profit or to fund additional charitable activities). Remember: cash flow is king (whether for-profit or not-for-profit).
I would suggest putting in parts orders for the second build of 100,000 before you get the first 10,000 built. I think you'll go through your first production run very quickly - it will get interesting in the second build and as you get a sense for the product mix (how many of each model you sell).
I can only wish you the best of luck. I'll be here with my Visa card when the first production models are available. Really!