Discussions on “On the number of modes required for statistical energy analysis-based calculations”

In Ref. [1], it was argued using two subsystems that "for the applicability of statistical energy analysis (SEA), what is important is the presence of large number of modal pairs in the interacting subsystems and not large number of modes in each subsystem". It was further claimed that "Even if there is only one mode present in a particular subsystem, if the interacting subsystem has several modes in the frequency band, SEA is applicable". The purpose of this discussion is to highlight that the conclusions made in Ref. [1] cannot be generalized to using SEA for the analysis of a system with more than two subsystems and such generalization might incur substantial errors.