A very interesting topic. I haven't read the book yet, but I surely will.
Now my 2 cents [hah! another metaphor
] on metaphor. I think we could see metaphor as a kind of scaffolding: it provides a general propositional framework for a particular reasoning or statement. So while the METAPHOR itself can not be assigned a truth value, the METAPHOR + CONTEXT can. Of course, the latter might still not be rigorously defined, but that's not particular to metaphors - it's the same for most of language.
As an example, let's take the use of the word "scaffolding" in the previous paragraph. Because you read this as a metaphor, you didn't think of the physical aspects of scaffolds, but rather of functional aspects of providing support, as a structure to work from on. However, the precise interpretation depends on what is 'worked on'. Moreover, I could have wanted to compare the physical nature of scaffolds to some other thing. So the metaphor in itself is an empty structure, that gets its content from the context.
However, there are some extremes. One is the case in which metaphors become so much used in a particular context, that they start losing their metaphorical qualities and develop into idiom. The context is no longer necessary for the interpretation, because the possible contexts are restricted to a very limited set. This, I think, is the case with 'No man is an island'. We know what it means; we can paraphrase it. Of course, it might marginally depend on the context (see the examples by Tengado), but it's meaning is more or less fixed. By the way, I don't think 'property is theft' is a good example, because I see that more as an ethical prescription, not as a proposition.
The other extreme is 'loose metaphor', poetic statements, where the combination of METAPHOR + CONTEXT might not result in a truth-evaluable statement. Here, metaphor is used so idiosyncratically, that we have no way of recovering the intended meaning, if there was one.
I think the most interesting cases lie between these two extremes. For instance, I can make up a particular metaphor that you've never heard before, and you'd still be able to understand it. When I say X is like Y, I'm not saying that X is Y. What I'm saying is that X shares shome characteristic with Y. However, this similarity might not very direct - and still you're able to pick it out of multiple possible, equally vague similarities. How do we achieve that? Why are some aspects more salient than others?
From a sociolinguistic and cognitive viewpoint, I think some metaphors are also more interesting than others. These are the metaphors that shape our thought; the archetypical patterns of explanation for certain phenomena. I don't think 'no man is an island' is such a metaphor - it's just a stock phrase, although it might be a surface aspect of a stronger, underlying metaphor.
What I find interesting is how a lot of our 'folk psychology', the way we understand other people, is based on a curious mixture of Freudian theory, modern psychology and psychobabble. We talk about 'repression of memories', 'neurotic', 'unconscious desires', 'accepting things as they are'. Presumably our everyday understanding of other people's minds is different from that in the 19th century.