Does the industry need a vendor-neutral synchronized multimedia format?

Over the last few years, I've become increasingly aware of the almost endless different ways of authoring and delivering synchronized multimedia presentations -- you know, the talking head with sync'd slides and a table of contents, transcript -- that type of thing.  Along with the varying different methods of implementing such a presentation, (which include various combinations of SMIL1.0, SMIL2.0 Flash, Quicktime, WindowsMedia with embedded events, WindowsMedia HTML+TIME, RealVideo with RMEvents, Java, Javascript, Screenwatch and others), there are countless desktop tools and web tools from dozens of vendors that all facilitate authoring and delivery of these kinds of presentations.  

Of course, one problem with this scenario is that content authored using one platform or toolset cannot be easily, if ever, ported to others, or updated to keep up with ever-changing client-side technology and competing de-facto "standards".  Of course, if the shelf-life of the content is short, this may not be a problem.  But if you need the material to be around for a while, or you need to publish to multiple standards and formats on your own terms (not your vendor's), or you want control and portability as a matter of principle and business strategy, you're apparently out of luck in today's marketplace.  I think this is why, even with a bazillion vendor solutions out there, there are so many homegrown solutions to this problem.  Companies want and need control over how their content is stored and delivered.

My question is this: is there a standard in the works anywhere that creates, essentially, a vendor-neutral and implementation-neutral way to represent the content of synchronized multimedia presentations.  A simple XML markup could represent the video and audio sources, slide sources and timings, and basic table-of-contents information.  It could also include descriptions of basic polling or feedback elements, as well as other kinds of presentation elements.  Basically, it would be an interchange format that various vendors' platforms would be able to import and export.  It would be simple enough to write by hand if you wanted to.  

I understand that SMIL 2.0 is capable of this and much more, but it's more of an authoring language with an explicit implementation in mind.  It's much deeper than just a content description language, although a subset of it might be appropriate for this purpose.   Other initiatives along these lines that I'm aware of include the W3C's Timed Text Format, various metadata efforts through ViDe, and more specific standardization efforts by groups like DigitalWell.org, and ISMA .

So...is there a demand for such a thing?  Is there one in the works?  If we built it, would anybody come?  Thoughts?