Because their patent basically covers ANY point to point videoconferencing connection, as if that wasn't something completely obvious to do.
In the world of unfiltered IPv6, it's quite obvious and easy. However, I can attest from personal experience that doing so from NATted endpoints (i.e. any phone on IPv4, most computers, etc.) is far from trivial -- if I recall, we ended up using some kind of corporate reflector server (expensive, since all data flows through it, leading to high connectivity costs) because that was the only way to make it work reliably. And even then it wasn't great, didn't help productivity any, and to this day we primarily use Email, phone, and chat for communications as the proven best tools for the job.
I haven't looked at the patent in question, but 10 years would put it well before IPv6 was deployed in any meaningful way. Setting up a circuit between two NATted devices without a reflector is indeed quite difficult and non-obvious.