Yup. And when you need to measure where artillery shells land, the GR model is overkill and you'll waste less time by using a Newtonian prediction. I think this proves my point?
No, because "less accurate but in practical terms the difference can often be ignored" is not the same as "the same accuracy".