Many of us have an overly simplistic view of orbital mechanics that fall well short of how it works in the real world. Take the example of India's failed Vikram lander [1]:

> The official said that predicting orbits below 50 km is very tricky and usually they are estimated with an accuracy of plus/minus 1 km. If the lander`s predicted height was at 30 km, the actual orbit height may remain anywhere between 29 to 31 km.

> Similar will be the error in horizontal direction. Such errors occur as local gravity varies widely due to large variation of density of mountains and plains.

Newton's gravity equations deal with bodies like the Moon as having uniform mass. That's not he case and it matters.

Fun fact: China's Three Gorges Dam slowed down the Earth's rotation by 0.06 microseconds per day [2]. That's actually not a big deal because the rotation has variance all of its own from tides, weather, etc.

So even if you had a density map of the Moon you'd still have to account for variance added by rotational variation.

[1]: https://www.zeebiz.com/india/news-on-indias-vikram-lander-di...

[2]: https://www.kinetica.co.uk/2014/03/27/chinese-dam-slows-down...