I agree that $5/month is a slightly hard sell right now, although there are certain parts of Eduset (such as assignments) that cannot be easily accomplished with a blog. In order to fix this, we are actually in the process of implementing every one of the features you described. Soon, students will be able to post content to their classes (if the teacher allows it), making classes more collaborative. Students will also be able to turn in homework and take quizzes on Eduset, and teachers will be able to grade that work.
As for pricing tiers, does anyone have thoughts on which of these two options would be better?
1) The current model, where all features are provided free for one class, while $5/month is charged to create an unlimited number of classes.
2) The model the commenter above describes, where certain features are made pay-only.
I'd look to offer something in the $30 range as your first paid tier.
What data do you have which supports this contention? Is it more of a "gut feeling"? I had that gut feeling once myself, and you know what? It is a lie. A myth. Totally contrary to empirically observable fact.
The average teacher in the United States spends $200 to $400 of her own money per year (varies with level and geography, and I've seen a few numbers quoted from different studies) on instructional supplies.
The average customer for Bingo Card Creator spends $32.50 on a program which most use once or twice per year. (Which is still cheaper than the laminated set of bingo cards they were about to buy at the teaching store. Those cards are on the shelves because they sell.) I have a trial-to-purchase conversion of about 2.3% for my online version, which is considered fairly good.
Some teachers do complain about the price. That is a good sign -- if no one is complaining that it is too expensive, you've priced it too low.
Selling software to educational institutions is largely a matter of playing Enterprise Sales. I don't suggest you make that choice lightly.
Along these lines, perhaps consider some way to allow those who use and love your product to bring others in with minimal friction - let them try out the product in an administrative and student role, without having to sign up and learn on their own. Most teachers I know won't try something out until they really see the value in it (i.e. students engaged, fulfills some kind of need).
As for your second suggestion, do you think a way for current users to send potential users an invitation to try a demo account (a sandbox) would do the job?