Abercrombieboy, what is your definition of 'works excellent'?
iChat AV with my iSight sends a video stream that is 352x288 pixel and roughly 20 frames per second. This takes quite some processing power to encode.
On top of that each received frame is blended into the next so that dropped received frames don't matter that much. This also takes processing power.
Apple's iChat AV is a very good video conferencing solution IMHO. At least considering that it doesn't require a T1 line to do decent smooth, rather hi-res video. Other video conferencing systems I've seen don't work as well. But they never claimed to. They either do fewer frames per seconds or a lower resolution (often compensated by doubling pixels) and they don't blend between received frames either.
In my book they would offer poor, choppy video but for these applications things would be considered 'working excellent' (as in 'never claimed to be a very good solution').
But perhaps I'm doing your application unjustice. Does it encode true 350x300 video in 20 frames per second, and smooth all received frames to compensate for dropped frames?