The strongest reason to treat the Popol Vuh as an old story rather than a colonial composition is that it can be watched being told before the Spanish arrived, in a medium no missionary touched. Classic Maya painted vases of the seventh to ninth centuries carry scene after scene that match it: a pair of young men shooting a great bird out of a fruit tree with blowguns; a bird deity with a ruined beak; twins confronting skeletal underworld lords; a figure emerging from a cracked turtle shell; ballplayers in underworld settings. Michael Coe set these correspondences out in the 1970s, and they have held up under four decades of further decipherment.
The evidence runs older still. The murals at San Bartolo in the Guatemalan lowlands, dated to around 100 BCE and so some sixteen centuries before Ximenez copied anything, show the maize god's death and rebirth and include a figure whom Karl Taube, William Saturno and David Stuart identify as one of the twins. Whatever those painters called the story, they were painting something structurally continuous with what the Kiche later wrote down.
Two qualifications belong with that. The correspondence is of scenes and figures rather than of a fixed text: names differ between Classic inscriptions and Kiche, the two twins do not always behave as the book has them behave, and the underworld lords do not line up one to one. And the Popol Vuh is a highland Kiche version, set down by particular lineages with a direct political interest in its closing genealogies. It is the best window available onto Maya mythology, not a Maya scripture, and there was no single Maya scripture for it to be.