Electron tomography is a powerful tool that is widely used in materials science to unravel the 3D morphology of functional nanostructures such as organic solar cells. However, analysis of the 3D information is limited by the data quality which is dominated by the beam sensitivity of specimens. Therefore, it is necessary to develop advanced acquisition schemes to obtain high-quality 3D morphological information. In this work we simulate the electron tomography workflow to study the influence of the electrons dose, tilt range, and tilt scheme on the reconstruction quality. We employ a slab model of rodlike filler particles in a polymer matrix that is directly compared with the simulated 3D reconstruction using statistical means and by characterizing edge intensity profiles. We find that using counterintuitive acquisition schemes, i.e., large tilt increments at low electron doses, the reconstruction quality is preserved or even improved. Furthermore, inaccuracies in determining the position of interfaces between materials are quantified in dependence of the acquisition scheme with implications for volume and connectivity determination. Finally, it is shown that for the chosen specimen geometry at a tilt range of +/- 75 degrees the resolution becomes isotropic regardless of the total electron dose and the tilt increment, which is a clear advantage for subsequent quantitative analysis.