In this article I am going to suggest that this arbitrary separation is meaningless.  Much like physics had to come to terms with wave-particle duality, biology must consider the same perspective where the answer depends very much on the question and how it is asked.

Clearly genes are the agents of change, however the gene-centric view of natural selection has several problems, not the least of which is that the gene can't determine whether it is selected for.  In fact, an important question that needs to be asked, is what does it mean to have a greater representation in a population?  Does a multicellular organism have a greater representation because of the number of cells it contains or does it only count if multiple organisms are involved?  If it is the latter, then what is the purpose in a multicellular organisms since that represents alot of work and little benefit to the gene?  More importantly, genes cannot act in any intentional fashion, nor does it make sense to consider them as having a sense of time or the future, so it begs the question to suggest that they exist simply to propagate.

Whatever the criteria for suitability may be within a species, it is at this level that sexual selection will determine what the compliment of genes is that will be carried by the next generation.  Individual modifications will establish what constitutes the next generation that will begin the cycle over again.  

However, it is equally clear that reproduction requires the existence of other compatible organisms, so any adaptation cannot out-pace the group with which it expects to interact.  In other words, being too successful individually will destroy the group and any possibility of future mates.  We already know that when a population threshold drops below a certain level, then regardless of individual traits, extinction is a likely result. Therefore natural selection must also enable group selection, at a minimum, as a limiting factor.

If we consider the analogy of a sports team, it becomes clear that each individual must be capable of performing to the best of their abilities, and yet ensure that their actions are synchronized with the requirements of the group or team. Any individual concerned only with their own objectives invariably disrupts the team and reduces the success of the group.

Similarly some organisms may benefit greatly by association with others of their kind, while others have minimal contact.  As the dependency on a group increases, so does the effect that natural selection will have on the members of such a group.  Traits that benefit the individual must also benefit the group for them to succeed.  This doesn't ensure the survival of the group, but it plays a role in determining how natural selection may operate.

Let's consider a hypothetical scenario for the evolution of the wolf.  If we begin with individuals only, it becomes clear that they are limited by size and speed to the prey that they can get.  Similarly if we consider their reproduction, then if we assume that they are not group oriented, then there must be some strategy to ensure the survival of their offspring.  This could be accommodated by having mating pairs stay together, or that the female be adapted to care for them on her own, etc.  Since most animals are not arbitrarily aggressive towards one another, let's assume that at some point two or more animals inadvertently participate in killing a larger prey animal.  Perhaps the animal is weak and presents an opportunity that can't be passed up and the amount of food is too great for a single animal to consume, so it may be easier to share than to fight over it.

It may be something as simple as this that gives rise to the idea of cooperating to gain regular kills.

In fact, perhaps the idea of cooperation is simply the result of a larger grouping of animals gathering for the purpose of mating.  Whatever the case, at some point it becomes clear that such a cooperative effort becomes an easier vehicle for getting food than hunting alone.  Once such a cooperative group forms, then the selection pressure on individuals must shift to include the preservation of the group in future generations.  Any individual born that doesn't have such a cooperative tendency will leave the group and pursue their own direction.  However, such an independent individual may, over time, find it difficult to find mates that are not also attached to groups, so their fitness will decline until they are no longer represented in the species.  While there is no question that the genes of each individual ultimately determine what kind of organism they are, it is difficult to argue that there isn't a corresponding selection pressure that preserves the group.

However, let's consider the alternative and assume that such cooperative social behavior never occurred.  At this point we could speculate that wolves might conceivably become larger, so that the size and strength was a beneficial trait which would result in individuals that have greater hunting prowess.  As the animals become larger, there is less benefit to grouping together and should they become large enough it may actually be detrimental since they might simply out-pace the available prey in a region.  This might explain why the very large predators tend to be less social and operate on their own more often.  

The previous discussion is simply a story and only intended to illustrate how a shift to cooperative group behaviors could occur so that group selection can become a factor in natural selection.  It should also provide an example of how natural selection may actually oppose the formation of groups, which is also a form of group selection, albeit in the negative.

It would seem that if we consider the survival of an individual cell, then a logical extension to group selection is when we consider the billions of cells that make up a multicellular organism.  Is this not group selection at the most fundamental level?  In fact, one could make the point that the genes in multicellular species are, by definition, operating on cell groups.

It seems that any meaningful interpretation of natural selection must include a range of influences and not confine itself to singular explanations.  While each perspective may be useful in examining a particular function, in the end it is as meaningless as arguing over whether something is a particle or a wave.