If you were launching satellites details like the exact radius of the Earth would be of interest, but we are talking about polarity (or even azimuth or elevation) that going to vary by numbers that are much smaller than the resolution of the instruments that are used to aim the dish. The worst-case scenario is that the radius number is off by 11km or 0.15%. The impact on the angle formula is a much smaller percentage yet.
If you think about it, this kind of variance could easily be undone by the sun passing from one side of the dish to the other and the thermal expansion of the materials changing where the dish is pointing.