What you can/should consider is buying a commodity 1U server
for $1000 or whatever sticking it in you rack and running Linux and
Moodle(s) on it. In real terms it has cost you almost nothing. When it
starts to get slow buy another and set up another moodle(s) for some
other chunk of the institution.Unfortunatly, this is not a procedure that will be acceptable at any major US institutions.
For one thing, spending tends to happen in discreet chunks at given times. In most large Us, moodle would be stuck on the slow hardware, faculty and students complaining about how slow open source software is the whole time, until the next funding cycle. The way most CIOs at big Us think is in 'packages', you buy a server
and a license, assign staff to install it, test it, and then you should
be able to pretty much leave it alone until the next semester break.
Also, IT staff can't just drop in-progress projects to set up and test a new server whenever the old one seems slow.
Thus, major upgrades (which most CIOs regard putting new servers online as) tend to be scheduled between semesters.
We generally do things like server maintenance late at night on the weekend to minimize student downtime, so this sort of ad hoc solution isn't going to be very popular with any but the most fanatical FOSS advocates.
Thus if there is no server size guide for Moodle, it makes it a hard sell to larger Universities (while at smaller ones the lack of skilled IT staff would likely make the CIO run screaming for the hills if he/she were told they may have to build a cluster sometime in the middle of a semester when the LMS seems to slow down).
This brings up a final major issue with your proposal: at Universities, the load tends to be greatest just at the end of the semester, when all the big mid terms are being given and all the students who have been slacking come racing back to try and salvage their grade. This is the very last time IT staff and admins want to try putting in some new hardware.
Open Source's future's so bright it burns... 
