For reply to comment see Kang et al., ibid., p. 535 (2003). The present author is unclear as to whether or not the authors' reply actually addresses the specific criticism made. Therefore, he offers the following observations in support of the initial comments. Consider the sound intensity probe axis to be oriented at 90deg to the axis of field symmetry, i.e. close and parallel to the plane of the aperture employed by the authors. In an ideal hemi-diffuse field, each of the assumed population of uncorrelated plane waves may be assumed to have a counterpart, the pair having wavenumber vectors of equal magnitude and lying in the same plane, with components normal to the axis of symmetry of equal magnitude and opposite sign. The pair contribute sound intensity vector components in the direction of the probe axis of equal magnitude and opposite sign. Each pair provides zero net intensity; thus that of the total field is also zero. According to theory, the angular distribution has a cosine form, as measured by the authors, and also by myself twenty years ago. This distribution is irrelevant to the analysis of diffuse field sound transmission through a panel. Each incident plane wave excites the panel into vibration uncorrelated with that induced by any other of the assumed population of incident waves. Hence, the vibration and associated radiation (transmission) of sound energy induced by the above-mentioned wave pair will add: cancellation does not occur