I think like anything, you need to define yourself a niche and attack it.
How are you going to differentiate yourself?
Remember, the basic functions of editing and playback are going to
in products from Real Networks, Adobe, Macromedia, etc. But who knows,
the big guys might trivially support it as an import/export format
just to say "I support another feature"
On the other hand, I don't think "just an SMIL editor" is a commercial
product, because it's too simple. It has to be combined with something,
like an existing audio/video authoring suite that is being converted
to support the web, etc.
If something is not complicated to implement, and easy to author
without a tool, it is probably a good candidate for freeware.
I was toying with the idea myself of implementing an SMIL player using the
Java Media Framework, but I've got too many projects on my lap already. :)
Now, I think a MathML editor/displayer has potential. :) For instance,
a tool that could convert LaTeX to MathML would be valuable, and
perhaps could export from MathML to Mathematica/Maple.