Or why Mark Gibbs is wrong when he says quality, support and integration issues hamper these 'closed' technologies.

Mark, interop happens when each party has an abiding interest in it.

Consider IM interop.

AOL, Microsoft and Yahoo only opened their messaging protocols part way. No interop on animation, file transfer, voice chat, webcam chats, conferencing, emoticons and variations on super-duper-emoticons, mood and presence messages, etc. Only the most basic of text message exchange.

Why the limitations? Because text interop is a commodity and the rest still have economic value. Interop for a commodity feature builds network effects without giving up capabilities that distinguish your community's identity.

Standards freeze innovation, for good or ill. VoIM (Voice over Instant Messaging like Skype) is still new. Skype is still evolving how to work in the hundreds of network and compute environments in hundreds of millions of situations. Those situations are challenging and evolving on their own. How you cross firewalls, scale p2p, and adapt streams for mobile conditions aren't ready to be locked down.

Like Microsoft, Skype opened up to third parties by publishing protocols. You can write your own voice mail client in an hour. Skype has APIs for presence, IM, voice and video calls, voice messaging, billing, and call routing. Skype has a budding developer community, mashup contests and free wrappers and other tools for programmers. Skype even publishes a high-level working API roadmap. It's still early, Skype's not done. And Skype is far from closed, Mark.

P.S. My first thought for a title was "Why Mark Gibbs must die" but I'm sure he's much too nice a guy for that sort of linkbait.