One thing this means is that kin selection gets classified as a form of genetic self-interest - rather than being classified as being "altruism". "Reciprocal altruism" is also classified as being selfish - despite its name.
However, others argue against such definitions, saying things like this:
As Sober and Wilson (1998) note, if one insists on saying that behaviours which evolve by kin selection / donor-recipient correlation are ‘really selfish’, one ends up reserving the word ‘altruistic’ for behaviours which cannot evolve by natural selection at all....and this:
If an altruistic behavior reduces the net fitness of the altruist and his kin, it cannot evolve.
The problem with these objections is that they are trivially wrong. Altruistic behaviours can be favoured by natural selection - even if they are deleterious to those displaying them. This is a simple consequence of the "extended phenotype" perspective - that traits may be coded for in different organisms from those that exhibit them.
Viruses cause all kinds of deleterious behavioural traits in their hosts - coughing, sneezing, itching, suicide, etc. However, these deleterious behaviours evidently do evolve via natural selection.
In the case of altruism, one of the the most likely cases involves manipulation of hosts by memes causing altruism:
Memes like to cause positive interactions between their hosts - since such friendly contact is one of the main ways they use to spread between their hosts. So, making their hosts eager to help each other can be expected to be the type of trait which many memes encourage. Such interactions may not necessarily be beneficial to their hosts (altruism is, by definition, a costly act). However, such behaviour most definitely can still evolve - via natural selection acting on memes.
I think that part of the reason we still see these kinds of explanations being offered is that understanding of memetics is not yet very widespread.