Zeuglodon, judging from comments such as #48 I'm not exactly making myself popular here, but I hope this forum can tolerate one more comment (or two) from my side.

It is because of the gene that a protein comes to be, but it is also because of the protein that a gene comes to be selected for. So I maintain that they can't be separated.

I never claimed to say anything revolutionary, since this is really just multi-level selection. Obviously I would never want to replace the gene level, I only argue that sometimes other levels can also be relevant. Of course genomes build on genes. I totally agree that with sexual reproduction, the importance of the genome level is greatly diminished, if not even irrelevant. I actually don't believe in parsimony in biology, since I have never seen any empirical evidence for such principles in biology. If anything, reality often turns out to be more complex than we think.

The mutation rate per amount of genetic material should be pretty much the same in genes and genomes (with some differences). Many more mutations in a genome, but taking place in a much larger material. If there are 2 mutations in a small viral genome and 20 in a much larger mammalian genome, do you then conclude that the mammalian genome is changing ten times faster...?

Now the r issue. You say:

The only way this is possible is if a new mutant gene arose that coded for this alternative strategy, in which case it would fail to dominate. Any genetic strategy has to work such that it can outdo a rival allele, and that means that kin-oriented alleles will prosper over universally-altruistic ones.

Also, In "Misunderstanding 5" in "12 Misunderstandings of Kin selection" RD says that the Kin gene will outcompete the Universalist gene.

But why could not the new U gene and the old K gene coexist, in different places in the same genome?

They could be producing two totally different proteins, but one could also imagine a gene duplication event followed but mutions in one of the genes. Either way, suppose that we have both genes, and that K influences our behaviour more strongly than U does. The result would be that we care mostly for our kin, but we also care to some degree about all other people. This would be a good description of what most people are actually like.

If U in combination with K has the effect of devoting some resources (not letting others starve as long as your familys survival is secured) to group members that are not close kin, but still actually shares most of the genetic material, the presence of U in the population should still lead to more copies of U and other genes surviving, all taken together.

The whole thing could also be epigenetically regulated. In a situation of plentiful food resources, universalist behaviour should be adaptive. Due to randomness in distribution, some people will always have insufficient food, but in a situation of plenty the cost for the haves to help the have-nots would be small. When the resources are scarce, it would be adaptive to be more selfish and save the resources for yourself and closer relatives, so as to not take the risk of having too little because of spending on the distantly related.

So lots of food could lead to U becoming more activated, and starvation to U being shut off. We know that starvation changes human epigenetic programming, so I don't see why not.

We also know now that there is not a single "intelligence gene" for example, but rather many genes all having different degrees of influence.

Why would altruism, another complex trait, be different and be regulated by a single gene that others cannot coexist with?