Seems like everybody is trying to ride on the IP network these days.  PBX's, video conferencing providers, VOIP.  Problem is, current networks are not designed for audio and video.  Real time conferencing components such as voice and video require significant "sustained" bandwidth.  Private networks have features such as QOS/bandwidth reservation, but not the public networks.  The next generation of IP (IP6) is better suited for rich communications but it may be some time before its widely adopted.  In the mean time customer will need to decide if these technologies are worth the bandwidth expense.  Microsoft Office Communicator offer audio and video conferencing out of the box, however peer to peer only.  Multi-party is supported with a partner such as Radvision or Polycom.  You will also need partner solutions to enable A/V over the public internet, since NAT and UDP do not traverse firewalls (research session border contollers/Jasomi).