There's no need to do that. Turning innovation into a process is exactly what happens at big, stodgy companies. They all are marching to the same drumbeat, innovating in exactly the same ways. It's all bland and boring - the really unique interesting ideas are an exception to the norm.
All those papers in the Stanford CS syllabus, guess who else has read them? Every other CS student that has ever went to Stanford. There are thousands of people who have had the exact same ideas you will have when you read those papers.
Just let your interests drive you. Let yourself fall down Wikipedia and reading holes. Traverse ancient web 1.0 sites with terrible design and a lack of meaning. Find a totally out of date computer science book and read it cover to cover. It's weird how the most dull, ancient history ends up being a wealth of ideas - just because no one has thought to look there yet.
If you want to create a profit generating product, then study up in some non-IT domain, understand the most pressing problems and solve one that current technology renders feasible. For example, the principles of generative text systems was conceived back in the 1950s. It wasn't until GPGPUs became sufficiently capable that the implementation of the theories became feasible.