I just took a few minutes to install a fresh copy of Drupal 4.6.5 along with the Aggregator2 module, which provides Glu functionality to Drupal. You can provide a bunch of feeds, and it sucks them in as if they are native content to that instance of Drupal. It tries to assign taxonomy/categories based on the data in the RSS feed, and provides some interfaces to view the feeds and items. I set up a prototype on my desktop box (so it might not be running all the time, and may disappear if my new box ever shows up), but it's there to play with. I've pre-seeded it with my blog, flickr, and del.icio.us feeds, and took the liberty of adding the same for another couple of folks (Brian and Alan) to see how it handles multiple users adding data.
My initial reaction is that it does the Glu stuff really well. But the query interface isn't nearly sophisticated enough to manage a full-on stream of hundreds (or thousands) of students pumping content into the system. Also, the Autotaxonomy module (the one that converts RSS categories into Drupal taxonomies) doesn't break the categories into individual items. If one post has categories of "eduglu prototype" and another has "prototype eduglu" - they are stored as separate distinct items in the taxonomy. It should break them into "eduglu" and "prototype" so I could find relevant items without forcing all users to use identical category strings. Which would fall apart if someone wants to add an additional category, like "eduglu prototype drupal" since it wouldn't match either of the other two literal string taxonomies.
It's possible that Drupal (and the Aggregator2 module) could be tweaked to do what EduGlu would require. Might be easier to do the work there, rather than starting from scratch...
Update: Interesting - Aggregator2 is properly understanding the individual categories on items published by my blog, but not from other tools like Flickr or del.icio.us - maybe Aggregator2 is right, and the others are wrong in how they're storing/sharing categories?