The reason this is such a profound mystery is that the initial cell somehow "knows" to divide and the daughter cells somehow "know" to form different kinds of cells which somehow "know" to migrate around the embryo and form different kinds of tissue which somehow "know" to integrate with other kinds of tissues to form organs, and so on.
So, how do cells with no brains "know" how to do all this? Where are the instructions located which choreograph this astonishing process and tell all the parts what to do and how to do it, and how are those instructions communicated?
The information is not to be found in either the genome or the epigenome, apparently, so where is it, what is its storage medium, and how is it stored and accessed? What mechanisms control it so that the entire assembly unfolds in a flawless sequence with each step occurring precisely when it must in order to successfully construct an adult organism? And how, exactly, does the zygote "know" to produce, say, a flower rather than a fish, or a bird, or a human?
These questions are fascinating and they emerge again in an article at Uncommon Descent that quotes geneticist Michael Denton:
The earliest events leading from the first division of the egg cell to the blastula stage in amphibians, reptiles and mammals are illustrated in figure 5.4 (in his book Evolution: A Theory in Crisis). Even to the untrained zoologist it is obvious that neither the blastula itself, nor the sequence of events that lead to its formation, is identical in any of the vertebrate classes shown.The blastula stage is an early step in embryogenesis when the zygote divides several times to produce a ball of cells. When those cells then evaginate and begin to take on the form of the early embryo biologists call that the gastrula stage.
Denton continues:
The differences become even more striking in the next major phase of embryo formation – gastrulation. This involves a complex sequence of cell movements whereby the cells of the blastula rearrange themselves, eventually resulting in the transformation of the blastula into the intricate folded form of the early embryo, or gastrula, which consists of three basic germ cell layers: the ectoderm, which gives rise to the skin and the nervous system; the mesoderm, which gives rise to muscle and skeletal tissues; and the endoderm, which gives rise to the lining of the alimentary tract as well as to the liver and pancreas....This is surprising. Different types of animals follow different pathways in building morphological structures such as the arm of a man, the foreleg of a horse, the wing of a bird, and the pectoral fin of a fish, that are otherwise believed to be evolutionarily "related."
In some ways the egg cell, blastula, and gastrula stages in the different vertebrate classes are so dissimilar that, were it not for the close resemblance in the basic body plan of all adult vertebrates, it seems unlikely that they would have been classed as belonging to the same phylum.
There is no question that, because of the great dissimilarity of the early stages of embryogenesis in the different vertebrate classes, organs and structures considered homologous in adult vertebrates cannot be traced back to homologous cells or regions in the earliest stages of embryogenesis. In other words, homologous structures are arrived at by different routes.
If they follow different pathways then there must be a different set of assembly instructions for the development of these "homologs," and thus all of the above questions arise again.
There is in the organism from the time it's just a single cell until it's fully developed, a massive amount of information that programs its development.
The locus, nature, and modus operandi of this information are unknown, but one thing I think can be inferred: If information of such astonishing sophistication controls the progression of the cell's development, it seems very unlikely that that information is the product of blind, impersonal, random processes.
Complex information such as we find in computer code or architectural blueprints are never the product of random processes like genetic mutation, but are always, insofar as we've ever experienced it, the product of a mind.
I leave it to the reader to draw his or her own conclusions. Meanwhile, here's a very good animation of human embryo formation and development: