The way I would describe the difference is really that Vidyo is doing what I would view as a SaaS model.  The licensing model seems to be on demand and the endpoints go "To The Cloud".  Then there is sharing within the cloud based on the call.

MCUs are very different.  Basically all endpoints dial one and a rented connection is established based on the call.

More efficient?  Maybe.  Depends on what you are doing. Let's take the following quote:

"By eliminating the need to connect all video endpoints to a centralized spot, the VidyoRouter Cloud Edition also keeps video traffic locally connected to a VidyoRouter, reducing the amount of transport needed and making the system more efficient and scalable, says Scott Morrison, research VP in the Enterprise Network Services group of Gartner Inc."

What??!!??  How does one keep all the video traffic local to a VidyoRouter if it is not all connected to a central point.

What I actually read was that the transrating/transcoding function (and for Craig I would say let's call transrating - degrading) is built in since all the endpoints have to start in a common codec and the protocol has built in degrading.

seven

I assume that the less efficient method of video conferencing described by Vidyo and the assorted analysts is the one used by Cisco. I wonder how Cisco feels about some upstart router vendor claiming it can corner the market by making a better case for service providers.

Polycom has a service provider angle, too, and they seem to have a pretty evolved view of the market. Their CEO says Cisco is silly for chasing the consumer telepresence space and they think that FaceTime from Apple actually helps them sell more systems because folks get used to the idea of video conferencing.

http://blogs.forbes.com/ericsavitz/2011/02/18/polycom-plots-move-into-consumer-video-conferencing/

Is Vidyo's technical advantage enough to make a dent in the space when Polycom seems to have a strong brand and the wind at their back?