Kahler manifolds, and RG flow in supersymmetric low-energy theories

I have a question about the relationship between Kahler geometry and anomalous dimensions in low-energy effective supersymmetric field theories. My question, in brief, is how (or if) the RG flow in the infrared is determined by the geometry of the metric on the moduli space. I have seen some authors derive RG flow as, roughly, derivatives of the metric with respect to the moduli. I have seen other authors claim emphatically that this is NOT what the low energy anomalous dimensions are. Suppose, starting from some renormalizable SUSY theory, we have a moduli space of vacua. When we flow to the IR, we have a generic nonrenormalizable Kahler potential. It can be shown that the action, up to superpotential, is a supersymmetric analog to a nonlinear sigma model. For example, the scalar kinetic terms take the form $$ \int d^4x \partial_\mu \phi^i \partial^\mu \phi^{*\bar{j}}g_{i\bar{j}}, $$ where $g_{i\bar{j}}$ is the Kahler metric on the space of vacua. Clearly, the metric behaves like a field strength normalization which varies along the manifold. In general, the action involves a bunch of nonpolynomial, analytic functions of the moduli. If we want a convergent polynomial action for the goldstones, we can pick a point on the manifold, and taylor expand. Then the deviation from the expansion point is our dynamical field, and the interactions are set by the coefficients of the Taylor expansion. The UV cutoff is set by the radius of convergence of the Taylor expansion. Since the UV cutoff is set by the radius of convergence, I could see the argument that changing the cutoff amounts to changing your coordinate on the moduli space, in which case the running is derived by taking partials of the coefficients in the expansion with respect to translation along the manifold (to greater or smaller radii from the origin of the moduli space, I guess) On the other hand I'm not sure if considering a vacuum with a smaller vev can really be considered equivalent to considering a larger fluctuation about the same vacuum. It seems to me that maybe then an RG transformation corresponds to a rescaling ie a Kahler-Weyl transformation of the manifold? Then the beta functions and anomalous dimensions will be the infinitesimal action of that transformation on the coefficients of the Taylor expansion of the low-energy EFT. Sort of a vague question I know, I'm just not quite sure how to reconcile all this.