Memetics vs multi-level selection

I don't really object to multi-level selection theory - but it does seem to be frequently over used.

For example, here's David Sloan Wilson explaining the need for group selection in explaining altruistic acts:

Early in the video, Wilson claims that group selection is the only possible explanation for altruistic acts. That's just wrong.

Checking with my list of viable hypotheses which can help to explain altruistic acts, it looks as though Wilson may be underestimating the influence of manipulation and overgeneralisation.

In manipulation, parasites and memetic symbionts manipulate their hosts for their own ends. Since symbiont reproduction requires contact between hosts, making the hosts be nice to other hosts helps the symbionts to reproduce.

In overgeneralisation, resource-limited cognition results in organisms being nice to one another - because that is what usually pays.

As with other group selection theorists, Wilson may also be over-estimating the actual incidence of altruistic acts. For example, many forms of altruism act as a form of virtue signalling - and show how nice you are.