> Does this mean instead of individual objects we make our calculations with x units of mass and energy in every given patch of space and this may be an issue,

> Does this mean instead of individual objects we make our calculations with x units of mass and energy in every given patch of space and this may be an issue, or am I getting this wrong? Pretty much yes, as we can only take observations like that. More importantly, the GR equations are non-linear, so taking these 'average' observations and feeding them into our models may introduce significant error. Or another way of saying that, it's possible that the model would give accurate predictions if you were able to feed it precise and complete information (the details of every single point mass) but not give accurate predictions if you feed it the sampled data we have (observations about whole galaxies, for example). People have discussed this elsewhere in this thread as the N-body problem. It's also saying that even if we did have that level of detail, our numerical simulations aren't currently capable of calculating at that fine a level of detail.