Well… Galileo was working with a heliocentric system that did have some flaws.
The geocentric model, from the Catholic perspective, had the benefits of being endorsed by Ptolemy and Aristotle. The view was simply that the stars are embedded in a crystal shell surrounding the Earth. Inside that crystal shell are a series of crystal spheres inside which are embedded the sun, the moon, and the focus of the planets’ epicycles. Ptolemy explained the retrograde motion of the planets with epicycles: the planets theoretically revolved on circles with a fixed point on the crystal shell at the center. It wasn’t perfect, but neither were the instruments of astronomical measurement at the time.
Anyway, the heliocentric Copernican system predicted the motions just as well as the Ptolemaic system, but not necessarily better. In order to make the predicted motions of the planets match the observed motions, it was necessary to add epicycles just like Ptolemy’s system. (Mathematicians didn’t stop with just one epicycle, either. Sometimes they added “epicycles on epicycles,” sometimes to the fifth or sixth degree.)
It wasn’t until Johannes Kepler finally concluded that planets don’t move in circles, but in ellipses with the sun at one focus that the geocentric view finally became obsolete. It could no longer predict the motions of the planets better than its competing system.
Kepler had access to Tycho Brahe’s highly accurate observations of the positions of the planets. Tycho’s observatory, Uranienborg, was a product of Renaissance ingenuity and was more accurate than anything Ptolemy or Aristotle had access to. Kepler had to develop mathematics that explained the observed motions, and the Copernican system (the one that Galileo was arguing disproved the Catholic dogma) was simply not up to the task.
The question of stellar parallax has been answered sufficiently.
The point is that Galileo was not operating on the certainty provided by Kepler. If the church challenged him to predict the motions of the planets within 2 minutes of arc, he would have given inaccurate predictions. (Mind you, so would anyone in the church. But still, what would that prove?)
It’s sort of like nowadays listening to Michio Kaku talk about how everyone should accept the superstring theory because it’s such an “elegant and beautiful” theory. Yes, but you can’t prove it, and you can’t use it to make predictions that differ from the Einsteinian physics we already have. For all we know, in fifty years, people will be wondering why no one would listen to Kaku because everything he said has turned out to be correct. And who knows, maybe in a hundred years people will be writing scathing e-books about critics of superstring theory and about the repressive political and religious environment in 2000.