Colonial observers did record the dark constellations. Cobo, Polo de Ondegardo and others noted that the native people saw figures in the black patches of the Milky Way and named them, and some listed the animals. What none of them recorded was that this was a system.
That is the difference between an observation and an understanding. The chroniclers treated the black shapes as one more superstition to be catalogued alongside the worship of stones, and having noted it they moved on. Nobody asked what order the animals rose in, or what the order was for, and so the calendrical function went unrecorded for three and a half centuries.
It was recovered by fieldwork rather than from documents. Gary Urton's work in the community of Misminay, near Cusco, in the 1970s established that the dark constellations were still recognised, still named, still ordered in a fixed sequence, and still used, and that the sequence corresponded to the agricultural and herding year. The Huarochiri material could then be read as describing a working system rather than an isolated belief.
This is a general lesson about the colonial record and it applies well beyond the Andes. A source can be accurate and still useless, if the person writing it did not know what question to ask. The gap here was not the facts but the structure holding them together.