efficiency when combined with the notion of "special creation" would predict that the human embryo would not develop features which would not be present in the child (gillslits and tails, for example), as doing so squanders energy and biological activity for no utility.
so, again, these ontogenic processes do *suggest* common descent.
That would be just your interpretation of what you think *special creation* would say. I don't see why *special creation* would be required, or even expected to say, that "the human embryo would not develop features which would not be present in the child". That would be working on the presumption that we know for a fact that there is no other purpose for having those features show up for a time and then disappear and I don't think science can say with certainty, that it has arrived at that point.
Just because there is no known reason for something now, doesn't mean that there isn't one and that it won't be found out sometime down the line. It just means that we don't know NOW.