I think you're attacking the question from the wrong angle. What will cause a project to fail is a poor understanding of the field, lack of funding, loss of interest/morale in a team. In other words, given an infinite supply of labor, time, and money, any project of any size can theoretically succeed. If you want to know more about what causes of failures for projects are, then search explicitly for that.
To narrowly define a game as "Art + Code" and ask which is more important is kind of like asking in the food service industry, "Which is more important in the success of a restaurant, food or spices?" I think you would be better off understanding all of the parts that go into a successful game development + launch and then trying to find points of failure in there.