I'm glad Harvard is willing to take on the task, forseeing OCLC's passing, but I'd rather see someone take on printing all the scanned stuff onto microfilm for permanent preservation (who is left with deep pockets and no pet project...can you say Facebook? I knew you could.) Intellectual access to something that has no physical access is useless, and it's scary how many institutions have lost files. (Can you say hot backup? Come on, try...)
In a related thread, the Chronicle has a fiery debate about Wikipedia. Personally, I'm fond of the person who assigned annotations as a class project, on the "better to light a candle than cuss" theory. I'd prefer the Stanford Encyclopedia of Philosophy model, but I can't afford a server either, so I"ll be happy with the guys who are doing the "portal" work for Wikipedia. It can work, it's just messy and slow, like life.
But the reality is that we have no national library (LC is not), no permanent agency, and no tax base (except NY state). There's a lot that KB can do because it has a big stick that's not available to us, so we need to accept that we're not going to be the leaders in this race.