We have recently introduced channeled-partial Mueller matrix polarimeters as a potential design for measuring a limited number of Mueller elements for remote sensing discrimination. Because in such systems the polarization information is modulated in space or spectrum, the corresponding carrier domain ends up sharing two different types of information, thus leading to a reduction of bandwidth for each. In this work, we concentrate on an efficient nine-channel/nine-reconstructables design, which limits the associated resolution loss by limiting the overall complexity of the system. Employing structured decomposition techniques allows us to produce a system description that provides an analytically deducible set of reconstructables that include o'š00, any two linear combinations of the elements within the diattenuation vector, any two linear combinations of the elements within the polarizance vector, as well as the linear combinations specified by the Kronecker product of the diattenuation and polarizance vectors. Finally, we optimize the available polarimeter parameters to align the nine reconstructables with the desirables derived from sample data, while maintaining the ability to discriminate between different objects.