Electron energy loss spectroscopy (EELS) is a powerful tool for understanding the chemical structure of materials down to the atomic level, but challenges remain in accurately and quantitatively modelling the response. We compare comprehensive theoretical density functional theory (DFT) calculations of 1s core-level EEL K-edge spectra of pure, B-doped and N-doped graphene with and without a core-hole to previously published atomic-resolution experimental electron microscopy data. The ground state approximation is found in this specific system to perform consistently better than the frozen core-hole approximation. The impact of including or excluding a core-hole on the resultant theoretical band structures, densities of states, electron densities and EEL spectra were all thoroughly examined and compared. It is concluded that the frozen core-hole approximation exaggerates the effects of the core-hole in graphene and should be discarded in favour of the ground state approximation. These results are interpreted as an indicator of the overriding need for theorists to embrace many-body effects in the pursuit of accuracy in theoretical spectroscopy instead of a system-tailored approach whose approximations are selected empirically.