Maybe worth pointing out that Abau is the only counterexample there to the ptk? rule - otherwise, everything lacking one of p, t or k has ?.
Yes, at least in austronesian small inventories often means that 'voiceless' vs 'voiced' is really just a naming convention - the voiced stops are often fricatives, nasals, prenasalised, or funky tongue stuff. Some people just call these series 'strong' and 'weak' stops.
[For a moment there I got confused, and thought that Iau was IaaI (ok, it's actually called Iaai, but that's less symmetrical). Surely, I thought, a language as insane as Iaai couldn't also be phonologically insane? But then I realised Iau was actually something from a long, long way from Iaai.
(But then I looked up the phonology of Iaai. Damn you, Iaai. You couldn't be sensible, could you? Ten phonemic vowel qualities, in two lengths, including both front rounded and back unrounded vowels. No plain labials, and instead series of palatalised labials and labiovelarised labials. Coronal, retroflex, prepalatal and velar stops series, where a stop series has voiced and voiceless stops, and nasals, and voiceless nasals. No labiovelarised labial voiceless stop... but also no palatalised labial voiced stop. Simple fricatives - one palatalised labial, one prepalatal, one velar, all voiceless... but then three coronal fricatives, including both a POA distinction not otherwise found (s vs T) AND a random voicing contrast between T and D, for no apparent reason. /h/, and a flapped rhotic, and then SIX APPROXIMANTS. Which six? Well, obviously, a lateral-alveolar (pronounced at the same POA as /s/ and a different one from all the other coronal sounds), a labiovelarised labial, and a palatalised labial. But that's only three? Ah yes, but of course Iaai has voiced and unvoiced approximants at all three POAs. But wait, isn't that seven? Because apparently /h/ is a glottal approximant in Iaai. Then again, on the other hand, the palatalised labial approximants (voiced and unvoiced) are apparently often turned into fricatives, although naturally these palatalised labial fricative-approximants, voiced and unvoiced, still contrast with the normal voiceless palatalised labial fricative, which is even more fricative.
Meanwhile, the retroflex and prepalatal stops are semi-affricated, but the coronal stops are very abrupt, and the coronal voiceless stop has an unusually 'short' VOA (this means it's more 'voiced', yes?).
Shorter version: languages with ten nasals shouldn't exist. They certainly shouldn't exist in the Pacific...)
But the river tripped on her by and by, lapping
as though her heart was brook: Why, why, why! Weh, O weh
I'se so silly to be flowing but I no canna stay!