I'll start with starting from the task. This is what is called popularly gamification, or what I previously decided to call game designed activity. I am not sure if I'll stick with that though. In an article I'm writing I'm using the term game-enhanced task. Moving back to the point, the core of this approach is that we have a task and by applying game design or game mechanics we seek to increase motivation in people to undertake that particular task. Scores and other means of virtualized feedback can be considered the basic case. The task is typically measured. We can introduce goals and challenges to improve the task's completion structure.
The other angle starts with the game. This is basically what serious games are about. Instead of adding enhancements onto a task, we take the task and include it as a key mechanic in a game. The key difference is that we re-frame the task entirely. Players playing the game need not even be aware of the actual task that is being done. To them it is simply an essential part of the game, and any benefit produced by the task is, again from their point of view, a side product. This bears a whole lot of resemblance to what is suggested by McGonigal (and many others I am sure) in her book and talks.
One interesting question is that do these approaches converge at some point? If we enhance a task enough with game-like elements, do we arrive at a point where there is so much additions around the task that has in fact become a core element of a game. Or do these two approaches start off in entirely different directions, resulting in applications that will generally not resemble each other. This is an interesting question, and one that I believe will be answered in the coming years when simple gamification techniques lose their novelty due to overuse. Soon the easy way will cease to work, then what?
Let's face it, the latter form is much harder to design. Some tasks are really hard to make into a game mechanic, even though it is possible to enhance them with game mechanics. Educational games seem to be running into this problem a lot: making an educational game that really teaches the subject while still being clearly a game is not an easy task. Mathematics has it easy: math is problem solving, games are problem solving. Framing mathematical problems as game puzzles is as straightforward as it gets. This doesn't make it easy of course, but easier. However, initially I think enhancing tasks with game mechanics will seem much more attractive, especially since the name gamification carries the illusion that it is easy (which it is not).
It is unavoidable that both of these approaches will face the same problems as games, or any products really. Novelty wears off quickly and after that only quality matters. However I believe that the future belongs to the game angle, at least where it is applicable. And it will become more applicable with ubiquitous computing technologies and smart game designers. However, all our beneficial games will have to compete with entertainment games (which is one problem with educational games). Are we up to task this time around? We shall see.
Anyhow, the implications of these two angles are an interesting question that I believe has not been asked yet. It is not a straightforward topic to explore, because whenever two approaches are combined there are a lot of variables and it is hard to control the ones that are not relevant to the study. This could well be my topic.