A source of confusion in discussions of ``fuzzy'' substructures is use of the same term for two types of structures with different characteristics. The fuzzy substructures discussed, e.g., by Strasberg and Feit [J. Acoust. Soc. Am. 99, 335--344 (1996)] have a multitude of closely spaced, deterministic frequencies of antiresonance and sufficient damping so that ``modal overlap'' exists over the frequency range of interest. On the other hand, Soize and other investigators consider substructures described statistically in terms of an ensemble of substructures, each member substructure having many antiresonance frequencies which do not necessarily have modal overlap, with their antiresonance frequencies and modal masses specified by probability distributions. Although the real, dissipative portions of the impedance determined for these two types of substructure are similar, the nondissipative, reactive portion may be different, resulting in significant differences in transient response. Moreover, the response determined for the ensemble of substructures may differ significantly from that of individual member substructures. Examples of the differences will be displayed.