Snacks and drugs both have direct, obvious 'value' to the body - food is about the simplest evolutionary case imaginable, and a host of animals will get drunk given the chance. And it's easy to make a case for 'information' as something any reinforcement learner will value and seek out, since it's essentially a prerequisite for all other learning and decision-making.
Money, though, is exclusively a proxy for some other future benefit, and lots of other examples imply we haven't made it into a strong first-order feedback loop. Money v happiness studies both show weak happiness benefits from money (once you're not in dire need), and find that "how you spend it" strongly mediates the effects. Slot machine design is devoted to the idea of intermittent reinforcement and works very hard to create non-monetary, directly limbic rewards to keep players engaged.
Given all that, "food, drugs, and info" actually sounds like a much more natural grouping to me than "food, drugs, and money". I don't know of any research here, though. Could you take a well-known study like prisoner's dilemma or the ultimatum game and get different results with non-monetary rewards, or prime people by discussing certain things they could do with the money?