One of the problems that gave rise to the gene-centric view is the prevalence of the notion that individuals tended to evolve for the "good of the species".  In essence the gene-centric view argued that natural selection operated for the "good of the individual" and specifically for the genes that would be capable of transmitting information to future generations.

Therefore if the genes are the information carriers, then they would represent the only means by which future generations could carry on from previous successes.

The problem is that it is meaningless to discuss individual genes in this context.  Using the analogy of a recipe for making a cake, this is tantamount to trying to determine whether it is the eggs, the flour, or the butter that are most important in producing the cake.

The point is that these are meaningless questions since the reductionist perspective is an absurdity.  It is only when the total configuration is present can we approach a meaningful consideration of what is taking place.

So it is with organisms.  They are not simply an arbitrary collection of genes that are striving to individually survive into future generations.  They are an arbitrary1 collection of genes that survives because, by good fortune, the organism survives.  As a result, it is the entire package which represents the components that can go into the next generation.  Is such survival the result of superior genetics?  It's impossible to tell.  The only thing that can be definitively said, is that whatever the genetics are, they were sufficient to ensure survival to the point of reproduction.

"Selection does increase the quantities of advantageous genes in a population, but this does not mean that individuals act to "maximize representation of their genes in the next generation. They act to maximize successful offspring __individuals__ in continuing, adaptable lineages. Increase of gene quantities is the result of this process; it is not selectively advantageous in itself".
Genes, individuals, and kin selection
As Dawkins states in The Selfish Gene, "The true 'purpose' of DNA is to survive, no more and no less".  Dawkins then goes on to state that non-coding DNA is simply a "parasite" that is "hitching a ride" along with other DNA.  

However this explanation is problematic, since any DNA that cannot participate in or contribute to natural selection also cannot assist in "maximizing" its representation in the gene pool.  This is the first problem with this perspective, since it suggests a degree of inefficiency at the most fundamental levels, so that on the one hand, genes are supposedly responsible for all manner of inherited traits to maximize their representation in future generations, while they promptly waste all manner of energy and effort to carry along these apparent "parasites".

Similarly, survival to reproduction, must extend beyond the individual into the species itself, since without a sufficiently large population and suitable mates, the genetic component is irrelevant.  Once again, the genetic material is mixed and the offspring will reflect a combination of genes that may or may not be successful.  Regardless of the nature of heritable traits, it simply isn't known whether each generation will be subject to the same environmental circumstances that allowed previous generations to survive.

What is clear is that whatever organism currently exists, it's ancestors had the good fortune to have the proper genetic mix and environmental circumstances to survive and reproduce.

Consider the example of the Peppered Moth which has become a common illustration of natural selection at work because of industrialization.  We find absolutely no evidence that the genes had any more influence over their final distribution of moth coloration than normal.  Instead we find that external factors (soot and pollution) determined which genes would tend to have greater representation.  When conditions changed, the original coloration also returned with increased representation.  The point here is that the genes were merely hapless "passengers" in the organism and demonstrated no capability of extending their frequency in the population as a result of their specific actions.

Specifically natural selection works because of diversity.  Having hundreds and thousands of choices to work from is what ensures that life will survive.  Therefore it follows that the most successful way to ensure the survival of any species (NOT the individual) is to ensure that the genes are mixed up frequently enough to avoid developing a "fatal flaw".  This must occur at the species level, since the survival of the individual is meaningless without a population group.

While natural selection cannot specifically select traits for the "good of the species", it achieves something quite close by forcing members of the same species to mix genetic materials to produce offspring.  This tends to "dilute" the effects of variations and avoids amplifying traits that could be deleterious.  This is amply demonstrated when species are isolated from one another and variations begin to occur.  If environmental circumstances are different between the groups, then the species traits will begin to diverge from one another.  Once again, illustrating that it is the environmental influence on the organism and not the genes that determine what information gets passed to future generations.  In other words, the information coded by the gene occurs well before selection occurs.  Therefore from the gene-centric perspective the future likelihood of propagation is no more certain than gazing into a crystal ball.  

Interestingly enough, I was struck by the following quote from an abstract.

"At the complex bio-system level, the genome context (the entire package of genes and their genomic physical relationship or genomic topology), not the individual genes, defines the system and serves as the principle selection platform for evolution."
http://www.ncbi.nlm.nih.gov/pubmed/19334004
Huh, did he just say the organism is the unit of selection?

1 My use of "arbitrary" is not to imply that the genes are merely a random collection of materials, but rather to argue that any particular genes from a particular species holds no more privileged position than any other genes from that species.