As to the name of the fallacy, I'm not sure. I suppose it's something like a misplaced expectation? The mistake is thinking that a certain theoretical moving part should do more work than it is rightly expected to do, while refusing to examine those moving parts which are rightly expected to do that work. EDIT: An example of a similar mistake might be thinking that a decision theory should tell you what to value and why, or that evolution should give an account of bio-genesis.

The SEP article's last section, on deontology and metaethics is very helpful here:

1. Deontological Theories and Metaethics

Deontological theories are normative theories. They do not presuppose any particular position on moral ontology or on moral epistemology. Presumably, a deontologist can be a moral realist of either the natural (moral properties are identical to natural properties) or nonnatural (moral properties are not themselves natural properties even if they are nonreductively related to natural properties) variety. Or a deontologist can be an expressivist, a constructivist, a transcendentalist, a conventionalist, or a Divine command theorist regarding the nature of morality. Likewise, a deontologist can claim that we know the content of deontological morality by direct intuition, by Kantian reflection on our normative situation, or by reaching reflective equilibrium between our particular moral judgments and the theories we construct to explain them (theories of intuitions).

Nonetheless, although deontological theories can be agnostic regarding metaethics, some metaethical accounts seem less hospitable than others to deontology. For example, the stock furniture of deontological normative ethics—rights, duties, permissions—fits uneasily in the realist-naturalist's corner of the metaethical universe. (Which is why many naturalists, if they are moral realists in their meta-ethics, are consequentialists in their ethics.) Nonnatural realism, conventionalism, transcendentalism, and Divine command seem more hospitable metaethical homes for deontology. (For example, the paradox of deontology above discussed may seem more tractable if morality is a matter of personal directives of a Supreme Commander to each of his human subordinates.) If these rough connections hold, then weaknesses with those metaethical accounts most hospitable to deontology will weaken deontology as a normative theory of action. Some deontologists have thus argued that these connections need not hold and that a naturalist-realist meta-ethics can ground a deontological ethics (Moore 2004).