Is Gould in the "Selfish Gene" camp, or has he broken out of that straightjacket?
(i.e. it limits evolution to a single strategy which is only suitable for zero sum games).

The argument for most genes being cooperative is much stronger. Individuals are produced by the (mostly) cooperative contribution of many genes. Natural selection normally operates on individuals, not their genes. Genes can't be the "fittest" by themselves (unless there was selective copying of genes) - they have to be expressed in a individual.

The concept of "fitness also applies to social groups, populations and whole species. i.e. a species which is well fitted to its ecological niche, but is not able to adapt to environmental change, is not the fittest in the long run.