An interesting case to look at for your question would be a child raised normally, except never punished for doing something we consider morally wrong, and never rewarded for doing something we consider right (and never observing others getting rewarded/punished). Obviously it's not possible to test this, though. My guess is that if you do that, you have someone who could commit horrible acts with no remorse, like a sociopath. Perhaps looking into famous sociopaths childhood would be interesting.
There is an Indifferent-Uninvolved model of parenting that fits some of the criterion you're supposing. In this model of parenting, as one could presume, the overarching theme is one where the parent acts in such a way that the child could not be considered an inconvenience. This is primarily accomplished through meeting a child's demand immediately, so as to terminate the child's demand, and by being emotionally unavailable to the child. Because child rearing begins at birth, the effects of such are immediate and consistent through a child's life, and outside sources of reward/punishment do not occur often enough to make a significant impact at earlier ages one could assume that this model of parenting loosely fits your inquisition. The documented results from such models of parenting happen to paint a picture remotely similar to the one you have. A truncated synopsis would imply that children coming from these households tend to be disobedient, impulsive, psychologically defunct (to some degree), and aggressive. This does not necessarily connotate to a sociopath, but it does share some of the same characteristics.
Interestingly enough, there is another model of parenting -which is widely lauded by mainstream child rearing manuals- that also fits the criterion you've proposed. This is called the harmonious model, and it contrasts sharply with what you seemingly believe to be the proper way to instill morals and direction in a child. The supposition that reward and punisment are integral qualifiers to instilling morality is thrown out of the window in this model. They are viewed as dated, archaic, and of relating to a hierarchy that is slowly but surely being debased. In this model, a parent will clearly outline expectations for a child, but will not necessarily enforce those expecations with punishment. Instead, this model focuses on creating a dialect between the parent and the child so that there can be a mutual understanding of why something is right or wrong. It also treats the child as an equal and does not generally profer superiority over a child when handling any given situation. The results of such child rearing are very telling of the misconceptions of reward and punishment in our society; a child raised under this upbringing will become self-confident, socially responsible, able to control emotions, and will generally have high self-esteem. The fact that this sort of parenting is seen as mollycoddling by believers in the strict reward/punishment model, indeed, says something about the actual validity of the reward/punishment model.
While neither of these exactly answer the question you have set forth, I believe that they can both give insight into the questions you have asked.
I think in the 60's some biologists were proposing that the process of natural selection might try to preserve larger gene pools in that manner but so far there's been no actual evidence of such a thing. Richard Dawkings proposed the single gene -view in his 70's book The Selfish Gene that has as one of its main thesis that you must always look at evolution from the point of view of a single gene, which became the standard view and has stood up to challenges so far.
This is a continuation of the conversation leading up to that quote, although what I am going to say is merely an insight of relevance to what has been said.
There is a certain type of organism (either an algae or a culture, I don't remember) that behaves in two distinct ways. This organism survives by floating on the top of water where it is capable of receiving nutrients and oxygen. It works together as a colony, but some organisms will self-replicate more hastily than others. The result of this, is that the entire organism will eventually grow too dense and sink, therefore drowning the entire colony.
This is a glimpse into how communal organisms may work for themselves as individuals (selfishly), while others work for the whole of the community. It is interesting because it brings to light the inherent dangers that can come from selfish behavior in a community, and it contradicts the statement that individuals form groups for the benefit of their selfish genes. Which, in this case, happens to be one of rapid self-replication.