A Reply to Makelov et al. (2023)'s "Interpretability Illusion" Arguments