Excellent idea but a bit unpractical because you can't stop being a programmer just like that - its the way you think. Also, I think that you would have to reach medium to expert levels in terms of skill/immersion by the time you know enough about the field to discover a problem worth your time.

As everyone else said, I would use existing skills to figure out what people think is wrong. Use that as a starting step to figure out what really is wrong. Then fix that.