Plural verbs, e.g. kikwah, have an underlying {h} plural marker. Thus kikwa 'he eats it' and kikwah 'they eat it.' The difference is clear when the verb is phrase final. kikwa has a final glottal stop, inserted automatically in phrase final position after final vowels. When the verb is followed by (e.g.) a noun, the {h} is lost in the surface form, but seems to break up the V-N phrase into two components, each with penultimate "accent". Thus in 67: {kikwa nakatl} > [kikwa nAkatl] one phrase group with penultimate "accent" {kikwah nakatl} > [kIkwa nAkatl] two phrase groups, each with penultimate "accent"
In the following examples, based both on the pitch tracks and on Jonathan's transcriptions, I believe that 187, 189 and 191 are singular, while the others are plural. 184 though 191 are the consecutive repetitions of this phrase given on the July 1998 Yale tape.
Overall, the observed F0 corresponds closely to Jonathan's description, with the possible exception of 187, where the overall F0 is lower than in the paired example 186, but the F0 relations among the first three syllables are not so different. Here in any case the F0 tracker is paying too much attention to small fluctuations at the syllable edges, as indicated by looking at the waveform shown at the end. However, the basic relationships are correctly determined.
I don't see any obvious indication of differences in duration or force of articulation. With such a small number of examples, even large differences would not be statistically significant, but in fact the pairwise segmental durations are within a few milliseconds of each other.
The only other thing worthy of note is that the female speaker shows clear downdrift in her "two-accent" cases (184, 185, 188), as does the male speaker at least in 190, though not so clearly in 186.
Waveform for example 187: