Nothing apparent in the sources, though the following observations from Bromley (1981) are probably worth a mention. The verbal paradigm of Dani is exceedingly rich in forms. For most of them, Bromley distinguishes between consonant- and vowel-final roots on the one hand, and between reflexive and non-reflexive on the other. Where the factive gerund is presented in the paradigm tables, it states (p. 360) that '[t]here is much variation in these rather rare forms, and the best informant often hesitated and was uncertain about the reflexive voice and vowel-final root forms.' This form for the verb 'come' seems to have been particularly problematic. (Not that the consonant-final non-reflexive forms seem to be unproblematic.) This seems at least to come close to defectiveness, though Bromley does not characterize it as such.
Van der Stap (1966), describing a related dialect, makes one observation which at least suggests the possibility of defectiveness. The medial voice forms involve what are apparently lexically-specified root (or stem) final consonant alternations; 'How the few verbs with root-morpheme-final n or p "behave" in this category is not known; the writer never heard them used; possibly they do not even exist...' (p. 84, fn. 37).
Bromley, H. Myron. 1981. A Grammar of Lower Grand Valley Dani [Pacific Linguistics C63]. Canberra: The Australian National University.
van der Stap, Petrus Antonius Maria. 1966. An outline of Dani morphology. 's-Gravenhage: Nijhoff.