First of all, I agree that I'm suspicious you're giving too much away in the first-page demo. That said:
You've priced based on a particular concept of what your customer is like: they're a large company with multiple employees carrying out process modeling. It looks like experience is showing you that concept is not correct.
You probably need to know more about your customers to really change the pricing, but I can take a few guesses at what's going on.
First of all, your $99 price point might be so low that an individual employee can put it on their company credit card without having to get purchase approval. If that's the case, what you've got is one individual at one company buying your product one time -- nobody's trying to use the cheapest version, they're all using the version they can most easily buy. Your model of how customers spend doesn't match how customers actually spend.
Second, you don't really know what specific needs differentiate one customer from another. What you need to do is position different feature sets of your product to match those different needs. (Adobe, for instance, tries to do this. Badly.) Number of seats is one possible need, but so is sharing, or output format (PDF download?), or different styles of model output (Rose, UML, etc.). You need to call up a number of actual or potential users and talk to them about the output they create and see how that differs. I bet that more big enterprise users need to share diagrams and edit them as a community, and probably also use styles like Rose, but I have no real data to make that bet. Call or at least survey your users and get that data yourself, then try to fit your users into 3-5 buckets of features, then price for each bucket.