I felt that the CS212 course by Peter Norvig was excellent -- the clarifications for the finals were so minor that I did not even notice them as odd, especially when we consider that they were multiple programming assignments to be done within a week. As time goes by, these things will only get much better (and will meet your criteria for serious value).

The course, in its present form, definitely beats the equivalent class at a large number of universities (and it will get better with each iteration). We cannot compare this with a similar, in-person course taught by Peter Norvig (though it gets tricky if such a course was in a big lecture hall with the condition that you lose if you snooze). It would be really unfair to compare it with the experience of a highly skilled and motivated group of friends taking a course at a top school.

Some major advantages I see are: (a) the lectures can be understood at one's own pace and are available 24x7 (contrast it with missing a class in college, or not paying attention in class for whatever reason), (b) there is a vast amount of discussion on course material (mostly junk but some material goes beyond how to get the right answer on a homework -- simply because of the vast number of people enrolled, and useful stuff gets modded up), (c) the format itself is highly interactive and engaging -- ironically, it is a more personalized experience (due to the large number of quizzes, ability to rewind, etc.,) than many classes in the university, and (d) this model can scale like crazy, not just with respect to number of students, but also with respect to number of courses [if we take a breather from the grading process, we can simply reuse an existing course].

Most of the information is out there on the web, but there is no systematic or guided process to learn. There is an Indian saying that goes like "knowledge without a guru (guide) is useless". These online universities provide such a virtual guide.

If the complaints about the courses are that some questions on the exams weren't correctly phrased and required clarifications... or that the professor had a math typo, then it is a good thing (since when did a typo became a sign of incompetence? he's not running for a close election in a polarized society).

... once you have saved to pay off the mortgage; the simple fact that you are no longer desperate to keep your job in a recession is worth more than all the toys combined.

In US, with rising property taxes, those taxes can be more than the rest of the mortgage payment, especially in some "good school districts". So, there is no solution to the desperation :). The definition of "own a house" becomes long term renting if one doesn't want to sell the house (or cannot find a buyer).

You need to specify additional information:

1) What about the data and storage? Many complex applications require vast amounts of data (e.g. climate change models, CFD models, GIS data sets that can complement or take advantage of modeling). Many end users may not be very adept at accessing these data.
2) What about the software? For example, CFD modeling software is very expensive. In some cases, open source software may not make the cut.
3) Does it have to be a single supercomputer? Why not split into multiple supercomputers and merge them as needed? That way, some groups have a more dedicated resource for themselves. The "biggest X ever" isn't as cool as it appears to be.
4) I presume the funds came in as a result of some proposal (using the word informally here, it could even be a one-pager that was sent to the university). The costs should be at least 5k per sever (based on what I've seen recently), so it's 6 million [I'd say 10 million even, unless my weak math is catching up with me]? So, that proposal would have some intended uses already.
5) Leasing it internally (to other groups in the university) may be reasonable -- it may even be a sweeter deal if you allocate a set of 10 or 20 servers for a group, instead of having it as part of a broader account access. You can tell them it is their "own machine".

I say this with no offense meant... I've noticed way too many people for whom the tool or technology seems to be the primary purpose (e.g. I do it using *EJB* or *distributed cluster* or *high availability database*). I spoke to someone that was working on app infrastructure for first responders, and was focusing on IDEs and integration, and his killer app was a download link to a weather channel app! When I mentioned that he needs some apps that really differentiate the system from others, his response was that we can run a contest for the apps. So, please avoid going that route -- in general, the tools are there to solve problems and not the other way round [with all caveats, sometimes the tools have to come first before we even realize what we had before that was very bad].

Well, congratulations on getting to play with 10M. I think I was rambling a bit, but the bottom line is: (a) don't make it one computer unless you can find a reason, and (b) approach different groups and offer the tool/service -- you need to do that till you get some traction.