Thats a great post and offers real insight, although a bit dated by ‘net standards, on one aspect of the Asterisk revolution that has failed to bring to the table a real player.
Asterisk is fairly painless to to get up and running, and with a little time and effort, to tweak out for a decent size deployment.
If your looking for clustering and imaging over VM’s using say, vmware (my weapon of choice) after the template is mapped out, custom kernel built and its screaming, your a step or two away.
Now rolling that into a slick interface with monitoring and fall back for the images is going to take a bit of doing. A single install with preconceived images that covers these bases, man, thats the ticket.
Now if you can imagine multiple ‘blade servers with this config, being deployed with PXE install, and you begin to see the advantage. Scalable, redundancy, and ease of deployment, you can’t ask for much more then that.
I think I feel a side project brewing :>