I'm a little biased, but I do think "voicexml" probably covered far more use cases than people are willing to admit. Some documents wouldn't have been pretty, I don't think, but would be far preferable to the inability to reason about how things actually work. And, indeed, the messier the document, the more obvious that you were making it too consuming.
Not that I know of? But, I have also not been watching that market for a long long time.
I am actually somewhat interested on building a small speaker thing that I can load a VXML document on and start adding use cases to it. Especially after considering how much I could save by dropping all of my current subscription stuff.
