Thank you for adding that info, Kent. That was very clear. It sounds like a *really* fun project. Like you, I was very disappointed when Adobe stopped supporting Pixel Bender -- that was the "poor man's" way for people to get their feet wet in this sort of programming.
Not having the original source code certainly ups the bar considerably on who is qualified to do this work. I can envision many different ways an algorithm like "Oil Paint" could be written, all producing slightly (or not so slightly, LOL) different results.
As I'm sure you know, there are quite a few pieces of software that claim to "turn your photo into an oil painting". IMHO, none of their results ever look like oil paintings, but some are interesting in their own right, and these companies already have demonstrated technical skills. Have you considered directly contacting these companies (often consisting of just a very small number of people), and proposing to them that you sponsor some work by them that might be mutually beneficial? For example, after a certain period during which you or your organization would have exclusive rights to the product that they develop for you under contract, they could then market it alongside their current products, and you share in the proceeds.
Anyway, let's sit back and give people time to respond because our members are scattered all over the world, in different time zones.
Tom M