We need a secular origin story, but is this the best we can do?
Darwin came up with his mechanism for evolution backwards, said his mentor Sedgwick. He came up with his theory first, then gathered evidence to support it. The result was a ramshackle bundle of assertions nicely skewered by Gertude Himmelfarb in her 1959 book "Darwin and the Darwinian Revolution," Part V.
Founders of the modern synthesis followed the same strategy. They too came up with assertions first, about genetics mainly, to provide them with entities they could apply mathematics to. They came up with random mutations, essentially random damage to existing genes, that by manifesting themselves in the phenotype would present targets for natural selection. They assumed genes contributed independently to the course of evolution and so their individual effects could be summed statistically, as the effects of atoms can be summed statistically to account for pressure in a gas.
The resulting "population statistics" became the basis for modern evolutionary theory. When evolutionists are challenged on some aspect of natural selection or mutation they typically claim their theory rests not on those concepts but on population statistics itself. But I believe those statistics were based on selection and mutation, by Ronald Singer for example, and if garbage went into the maths through flaws in the assumptions behind it then we can expect garbage to come out. And in books like Pigliucci and Kaplan's book "Making Sense of Evolution," even though their intent is to prop up the modern synthesis, garbage does seem to be all that’s come out.
So I regard the modern synthesis as not theory based on science, but an origin story based on mythical entities. Then the most appropriate questions are, what natural philosophy is being supported by the story, and what is its relation to our other origin stories?
While he was speculating about evolution Darwin read a review of Auguste Comte's "Positive Philosophy," one in a series of books Comte wrote in support of his reductionist thesis, that all scientific knowledge should be based on sciences at a more fundamental level. According to John Stuart Mill's account of Positivism, scientists were to omit any consideration of "volition, natural or supernatural." This omission of human volition might be essential for experiments to repeat precisely each time, but the effect could be to exclude human volition even when it was a vital part of the object of study, as in the case of human evolution it certainly was. By limiting himself to only purely physical mechanisms, in accordance with Positivist principles, Darwin illegitimately excluded human consciousness, creativity and free will from his tally of what a theory of evolution had to account for.
Scientific acceptance of Darwin's purely physical mechanism ended up denying any physical reality to human mental powers. This shielded Positivism from having to account for those powers, making Darwin's mechanism of natural selection very precious to scientists basing their work on Positivist principles. Taken as a myth, Darwin's theory supports Positivism and the natural philosophy based on it, physicalism.
Of course, this put Darwin's theory in opposition to Christianity, for which the mental powers Darwin omitted were vital components of the soul. Creationism is an origin story created to justify belief in the soul, in which God functions as the primary supporting myth.
This raises the possibility that other myths might be created to support other natural philosophies. I'm engaged in coming up with a theory of evolution able to account for consciousness, creativity and free will to serve the natural philosophy scientists like to call "folk psychology." Google "darwin vs galileo" for a video sample.
Appendix: what's mythical in modern evolutionary theory?
Natural selection implied a continuous source of variation, to replace the variations selection eliminates as it operates. By taking such a source for granted, Darwin was in effect calling on a mythical agent such as the Ancient Muses, able to prompt creation of variations at will, out of nothing.
For this source of variation population statistics would later conjure up the concept of random mutations to genes. No other source of variation now need be considered, any genetic variation found could be assumed to originate through random mutation. In fact, though, even if mutations exist they do not survive into the phenotype at random; the results of gene duplication undergo an extremely powerful repair process, so what is presented to natural selection is no more than whichever particular kinds of damage survive the repair process. And some genes are thought to be so essential they are conserved, protected from mutation; if so, then clearly mutation is not all random. If some of it is not random, perhaps none of it is.
To make mutations suitable for feeding into population statistics equations, they are divided up into categories: lethal, harmful, neutral and beneficial. For those equations, lethal and neutral are ignored—they are included only to give an air of authenticity. Harmful mutations are those that reduce fitness, beneficial mutations are those that increase fitness. Both survive in the gene pool, subject only to the effects of natural selection. Equations can now be written that sum up the effects, generation by generation, of the effects of natural selection on beneficial and harmful mutations. It is granted that random damage to the specs for a very complex living creature are much more likely to be harmful than beneficial. So in the absence of a beneficial mutation the much greater number of harmful mutations are bound to accumulate generation by generation, thwarted only slightly by natural selection, leading to a rapid increase in harm and speedy extinction.
Particularly mythical is the beneficial mutation. Such mutations will be favored by natural selection in every generation, so there is a similarly slight tendency for them to increase in incidence over many generations, until eventually they become the primary allele at their location. However, since this would take very much longer than it would take the harmful genes to drive the species into extinction, presence of a beneficial mutation must somehow banish all those harmful mutations, or render them impotent—the mechanism is not specified.
Beneficial mutations are as mythical as unicorns. If a change to a gene appears, whether on not it is beneficial cannot be established until millions of years have passed (except for micro mutations in labs, that I discount). Suppose a gene for salt tolerance is identified in elephants, is that a favorable mutation? Or neutral? Or is it noise in the system, amounting to harm? Only a time traveler could tell—do elephants take to the sea in ten million years? The concept of a mutation being beneficial is a myth, similar to the square root of minus one, created solely for insertion into an equation.
The use of myth in the modern synthesis has the same function, I believe, as natural selection alone—to support physicalism. Physicalism as a belief system requires that our evolution employ only purely physical processes. Since we’re probably several hundred years from coming up with the concepts necessary for truly understanding evolution, myth is the best we can do, and any myth that can be concealed behind a wall of impenetrable statistics can serve the cause of physicalism, for the time being.