This is meant to tell us what the unit of selection is, but !!!!
At the end of chapter 5, he says "we can ... treat the entire genome of an asexual organism as a replicator." Bearing in mind that this chapter is meant to be all about MINIMAL units of selection, this is a huge claim!
Options for how that isn't stupid:
1. Maybe it IS stupid.
1a. Maybe he wants a separate but similar theory for asexual organisms, but hasn't got around to it.
2. Maybe the theory of the whole chapter actually does work for asexual organisms, more or less.
2b. Can look at the largest pieces of a genome that are coherent and relatively unchanging.
(Transposable elements make it hard to define anyyhing as coherent and relatively unchang, tho)
A problem with 2b is that he keeps contradicting it, both a few times in the details of the chapter and again at the end of the chapter when he says you can "treat the entire genome of an asexual organism as a replicator."
3. Maybe the correct theory for asexual organisms is so simple that we don't have to worry about it.
We really don't believe 3. For example, if you treat the whole genome as the (only) replicator then you can't talk about details in terms of natural selection at all, e.g. horizontal gene transfer can't count as natural selection. And asexual organisms even have recombination.
Do we believe there's really a germ-line/dead-end distinction?
If we get rid of the germ-line/dead-end distinction but we keep the active/passive distinction, is what's left of Dawkins's theory still just as good? We want to be able to say that worker bees' genes are active (because they really do contribute to their own probability of reproduction). Well, we can just say that.
By the way, Dawkins says worker bees' genes are dead-end ... BUT ... he only does this by referring to them as the "chemical entities" (philosophers would say the TOKENS) rather than as "information" (philosophers would say the TYPES), but the rest of the time he's totally talking about the information/types, and rightly so, because that's what natural selection is all about. If (hypothetically, because we disagree with this) genes were chemical entities/tokens, then we'd have to agree with Dawkins's germ-line/dead-end distinction. But we don't (and I'm sure we're right about this bit), so we don't.
Side issue when talking about whether Dawkins's theory is (according to some) a bit rubbish. I'm used to treating "false" as not so bad, and "a bit rubbish" as much much worse. I wouldn't call something a bit rubbish if it's useful (but I would very happily call it *false"). Some of us disagree with this.
In some very important places - e.g. in the definition of alleles - Dawkins slips in an assumption that we're only talking about diploid organisms. But roughly 0% of all organisms are diploid. So !!!!!!!!!!!