Across species, neurons possess numerous ion channel types, with a single cell typically expressing tens of channel genes, which translate to hundreds or even thousands of channel protein types, each with different kinetics [1]. Moreover, maximal conductance densities across individuals are also highly variable and subject to neuromodulators which are essential for enabling the nervous system to switch between behaviorally relevant modes [2,3]. This raises the question of how neuromodulators can reliably induce changes in intrinsic neuronal properties across a heterogeneous population. At the same time, neurons must navigate metabolic constraints imposed by homeostatic mechanisms. How can neurons simultaneously satisfy homeostatic constraints while modulating activity in a high-dimensional parametric space of conductances? We propose that redundant ion channel expression facilitates these modulatory transitions while enabling homeostatic maintenance of important physiological variables, such as calcium concentration [4]. By simulating biologically relevant switches in the firing patterns of conductance-based model neurons of the stomatogastric ganglion we show that modulation of dynamic features is possible using a simple Gauss-Newton update rule for conductances. We use intrinsic neuronal feedback gains [5] to extract excitability features of interest and demonstrate how the sensitivities to specific features of neural function can continuously tune the electrical output by inducing coordinated changes in maximal conductances. In our framework, modulation is cast as an adaptive control problem where a few intrinsic features are tracked by the neuron's dynamics. We show that homeostatic modulation of conductances yields trajectories with less curvature, defined as the ratio of path-length to euclidean distance, when considering a larger, redundant cohort of ion channels. This is in agreement with the monotonic dose-response action of modulators, and corresponds to simple and plausible pathways for controlling neural excitability. Our work therefore offers an explanation for the profusion of ion channel species that many types of neurons express, and provides a principled method for controlling conductance based models.