Unfortunately I'm not a patent attorney and am thus out of my depth regarding actual patent law, but a friend and I were just discussing the concept of "obviousness". To my understanding, in order for an idea or process to be patentable it can't be "obvious" to a typical person in the field -- it has to be "novel". In my recent discussion, my position was that an idea is almost guaranteed to be novel if no one is doing it and it turns out to be profitable. Profitability plus lack of action implies that no one has thought of the idea.
My friend was arguing that there are some ideas that are so incredibly obvious that no one writes about them, that no one thinks will be profitable, and that actually are profitable once someone does them -- but I don't think that's the case. There may be some such ideas, but considering that people are wont to try vast numbers of ideas on the mere suspicion that they might be profitable, it's hard to believe that there are many obvious, profitable ideas that everyone thinks will be unprofitable.
Thus, my argument was that no matter how obvious an idea may seem in hindsight, if no one has done it or written about it, and it turns out to be profitable, then the idea is almost certain to be novel.