I am currently reading Dangerous Words: Talking About God in an Age of Fundamentalism by Gary Eberle, a literature professor at Aquinas College in Grand Rapids, MI. While there is much to promote in this book, I think the attempt to resurrect the language of myth and metaphor as positive aspects of our general human tendency toward God-talk is ill-advised and ultimately fatal to any serious attempt at discussing religion. Here's why.
I agree with his general historical discussion of the way "myth" was reduced to "falsehood" over the course of the 18th and 19th centuries, as especially English sought to purify itself of equivocity and sought a scientific-like clarity. In the course of doing so, it dismissed "myth" as pre-scientific language that sought to do poorly what science does well. This general approach transcends English, as the German New Testament scholar Rudolf Bultmann built an entire career out of attempting to demythologize Christianity - strip it of its unscientific pretensions to knowledge - and transform it from metaphysical language to existential language. Bultmann's former colleague, Paul Tillich, and Tillich's new American colleague, Reinhold Niebuhr, continued Bultmann's work here in the US. In the process they accepted uncritically the idea that "myth" is "false", whereas other, demythologized language is "true" or perhaps even "True". In so doing, they muddied their own work, and utterly destroyed the notion that mythic language conveys meaning. It is almost impossible to resurrect the idea of myth as meaningful discourse, because all of us, no matter how hard we try to deny it, accept that it is "false".
Tying "myth" to "metaphor" is usually a way to resurrect mythic language. The problem with metaphor, and not just in this general sense but even in very specific instances such as Sallie McFague's Metaphorical Theology, is that metaphors are inherently circular - words referring to other words that refer to other words, with the implicit notion that we are only drawing and redrawing lines on a conceptual map without referencing anything intersubjective or socially accessible. When taken to a usual extreme, as Eberle does, by saying that all religious language is metaphorical, we end up with two problems. First, we are saying that religious language has no referent; and we say that the specific differences between religions and religious vocabularies are conventions to be dispensed with. The former I find useless; the latter I find condescending.
It is all well and good to speak of "the human religious impulse" and "the human religious phenomena". It is another thing altogether, however, to say that the different human religions are essentially the same, just dressed up in different language. By removing specific difference in this way, we no longer have to go through the painful, wrenching process of actually learning about other religious beliefs, forcing ourselves to question our own in light of the explicit differentness and otherness these religions pose. All we have to say is, "You're doing the same thing I am, only you go about it a little differently." Real dialogue, real understanding is short-circuited by the paternalism of essentialism.
The specifics of Christian God-talk have to be taken on their own terms, not as metaphor, not as myth, but as real human struggles of real human communities to make sense over time of their experience of what they have come to experience as God. It is one thing to insist that, being contingent, fallible, and limited, human language about God is necessarily flawed and imperfect. It is another thing altogether to say that it is metaphorical, mythic, and essentially the same as all other religious language in all other traditions. It isn't, and it is both ignorant and paternalistic to claim it is so.