I think the problem is that too many people treat VXML like it's a gold mine or somehow valuable.That is a no no for a standard.
So when it was first invented, people tried to sell the editors for money, the only implementation were commercial and there was no real acceptance to it. The first open source VXML stack was dependency-ridden and archaic to even build for many years.
The value could be realized by providing hosting services for it, or developing it for customers, not by making them pay money to even use the editor.
I believe a standardized way to serialize an IVR and execute it from a neutral server is a good idea but I am a bit sketchy on whether or not VXML is right. Imagine if email was a secret protocol and you had to pay money for your email composer...Oh yeah, that happened and everyone laughed in their faces and now it's free.
When I once tried to implement VXML, I started trying to use openVXI, while hunting down the dependancies, I learned about spidermonkey and started to wonder why to bother rendering the XML into a script and running it when you could just write the whole IVR in JS. So I made res_js for Asterisk. When I started FS, I also made a mod_spidermonkey as we can all see. I also use XML in FS so I don't have a vendetta against XML, (I do get annoyed by those who over do it with XML). I just get an uneasy feeling about it and there is something to be said since we have 100% of the ingredients for VXML in FS and never implemented it.
Like I said I am open to the idea of supporting it, to give it a fighting chance to get adopted. But teenagers are harder to get adopted as Andy pointed out =p