Neo-Riemannian theory is a very broad subfield of music theory, but we can generalize and say that it is an associative theory. In other words, it is a network-based theory that relates how a given entity relates to other entities. (You addressed this, you just called it "interrelational.")
Schenkerian theory, on the other hand, is a hierarchical theory. It is a theory of monotonality, which means that there is a single tonic that governs a given span of music. Thus it is a theory that relates how all given entities relate to a single governing entity. (You called this "hyper-tonal.")
Now we can look at the distinction between rhythm and meter. Whereas rhythm is a pattern of articulations, meter is the hierarchy of that pattern. Thus a study of meter is necessarily hierarchical, and so an associative system like neo-Riemannian theory would not be ideal to study rhythm. But again, you implicitly address this, because you specifically asked for approaches to rhythm, not to meter. (You did your homework!)
But the problem is that music theorists often join rhythm and meter in their analysis, so it's tough to find analyses of only rhythm (the associative connections), with no concern for the meter (the hierarchy of it all).
Nevertheless, you may want to check out: