Recently we solved, via discrete numerical grids, the Ramakrishna-Yossouff density-functional theory equations for the freezing transition and obtained an intricate phase diagram of hard-sphere mixtures. Even though such methods provide more variational freedom than basis-set methods, we found that the thermodynamic quantities were sensitive to the spacing of numerical grids employed and observed numerically induced false minima. Dasgupta and Valls have commented that these false minima were due to our use of k-space methods and, hence, their early works based on a fully r-space approach are qualitatively correct, despite also being sensitive to the mesh granularity. Here, we clarify the issues of achieving correct thermodynamic limit from grid-based methods and respond to their Comment, concluding that r-space calculations using coarse meshes may provide correct thermodynamic quantities (only by extrapolation) and thus their previous work should be called into question. In general, both methods, k-space or r-space, suffer from grid-induced problems.