Excellent book. I think group selection is a more plausible mechanism towards the creation of gradually more complex forms of life than kin selection, although I think the latter may be important to explain a basic form of institution in humans, the nuclear family and the love we express to our close relatives.
However, I disagree with the point of view that individual selection is always opposed to group selection, and that group selection is the only force that drives a group toward eusociality and complexity. It sure is the case for altruist behavior, but complex biological systems are formed by other types of relations that could be selfish AND also confer an advantage to the group. Take specialization in labor division in humans. Many people study a career to make money (a selfish motivation), but specialization is good for the group because it saves energy and resources to that group. In this case, the group does not have to be in competition with other groups in order for the trait to evolve. It evolves because it is advantageous in the individual level too.
But specialized individuals are simpler than non specialized individuals. Specialized individuals have to perform just one task, but non specialized individuals have to be able to perform many tasks. Complexity rises when individuals specialize because interdependance increases, increasing with it the number of connections in the group network. Simplification in nature is a more probable output when a mutation takes place because it increases entropy (although simplification can be attained by epigenetic modifications). Then, simplification (and further specialization of individuals within the group) at the individual level is paired with an increment in complexity in the group level.
As we can see, the growth in complexity based on this theory does not need group selection at all. It is driven by internal forces within the group.
I think that complexity emerges from a combination of these mechanisms: group selection in one side, and increasing division of labor without external influences in the other.
So, further complexity in human societies does not necesarily imply further evolution toward transforming ourselves in better human beings with our fellows, or being more aggresive with external groups. Further specialization and technology by themselves can do the job. Consider internet, a big leap toward a global human superorganism.