Maya astronomy was accurate, sustained and written down, and it was aimed at prediction rather than at explanation. The Dresden Codex contains a Venus table tracking the planet's synodic cycle over centuries with a correction mechanism, and an eclipse table giving the intervals at which a solar eclipse is possible. Neither is a model of what the planets are; both work.
Venus mattered most. Its appearances as morning star were tied to warfare and to the taking of captives, and inscriptions record campaigns timed to them. That is astronomy as statecraft, and the tables were maintained because getting the date wrong had consequences.
The sky figures themselves are less well recovered than the tables. Tzab is secure, the peccaries at Gemini and a turtle at Orion's belt are reasonably so, and beyond that the reconstruction is thin, partly because the codices were burned. Three or four survive out of what must have been a substantial library, and Diego de Landa, who burned them at Mani in 1562, also wrote the account that made much of the script decipherable four centuries later.
The atlas records the figures that are secure and does not fill the gaps. A Maya sky reconstructed to look as complete as the Greek one would be mostly invention.