Coincidentally, I'm looking for a python coder to work with on this next one. If anyone is interested in learning more, send me an email.
The project is using a Django extension called Piston: https://bitbucket.org/jespern/django-piston/wiki/Home
Which is really quite nice for building REST APIs with Django.
(From their company page, http://mashery.com/company: "We help brands... manage (their) API Powered Platforms..."
And for high-level guidance, http://redmonk.com/, kind of a Gartner for platforms/developers.
The kind of clients that want to build an API 1) already have a main product/website and 2) are likely more technical than average. It's probably something they could do themselves.
I write APIs. I specialize in them, but for a specific purpose. I develop 3d animation technologies and then expose them via an API. I then license use of that API to 3rd parties, hosting the servers in both cloud and colocation. That is something tangible they can grasp. This is after spinning my wheels for a while trying to do what you seem to be thinking.
It strikes me that quite possibly one of the next channels that businesses might want to exploit are having their data available programmatically. Which is why I'm wondering whether that's one of the next market opportunities.
While its not our core business, we host content APIS for some large publishers. Because we're ingesting their content anyway (content syndication is our main business) and we already have a robust API, we've productized it to provide APIs as a service back to our content providers.
I think content and ecommerce APIs are where you will see the most traction.
Following the same train of thought, why aren't there shops out there that ONLY do "web forms" (not talking about SaaS here), or webdesign agencies that only do "landing pages"?