Most density functionals lack to correctly account for long-range London dispersion interactions, and numerous a posteriori correction schemes have been proposed in recent years. In van der Waals structures, the interlayer distance controls the proximity effect on the electronic structure, and the interlayer interaction energy indicates the possibility to mechanically exfoliate a layered material. For upcoming twisted van der Waals heterostructures, a reliable but efficient and scalable theoretical scheme to correctly predict the interlayer distance is required. Therefore, the performance of a series of popular London dispersion corrections combined with computationally affordable density functionals is validated. As reference data, the experimental interlayer distance of layered bulk materials is used, and corresponding interlayer interaction energies are calculated using the random phase approximation. We demonstrate that the SCAN-rVV10 and PBE-rVV10L functionals predict interlayer interaction energies and interlayer distances of the studied layered systems within the range of the defined error limits of 10 meV per atom and 0.12 Å, respectively. Semi-empirical and empirical dispersion-corrected functionals show significantly larger error bars, with PBE+dDsC performing best with comparable quality of geometries, but with higher interlayer interaction energy error limits of ≈20 meV per atom.