Molecular photoswitches can under certain conditions be used to store solar energy in the so-called molecular solar thermal storage systems, which is an interesting technology for renewable energy solutions. The current investigations focus on the performance of seven different density functional theory (DFT) methods (B3LYP, CAM-B3LYP, PBE0, M06-2X, ωB97X-D, B2PLYP, and PBE0DH) when predicting geometries and thermochemical properties of the [2.2.2]-bicyclooctadiene (BOD) photoswitch. We find that all of the investigated DFT methods provide geometries that are in good agreement with those obtained using coupled cluster singles and doubles (CCSD) calculations. The dependence on the employed basis set is not large when predicting geometries. With respect to the thermochemical properties, we find that the M06-2X, CAM-B3LYP, PBE0, and ωB97X-D functionals all predict thermochemical properties that are in good agreement with the results of the CCSD, the CCSD including perturbative triples (CCSD(T)), and the explicitly correlated CCSD-F12 and CCSD(T)-F12 models. Lastly, for energy calculations, we tested the newly developed fourth-order cluster perturbation theory singles and doubles CPS(D-4) model, which in this study provides energy differences that are of CCSD and sometimes also CCSD(T) quality at a relatively low cost. We find that the CPS(D-4) model is an excellent choice for further investigation of BOD derivatives because accurate energies can be obtained routinely using this methodology. From the results, we also note that the predicted storage energies and storage energy densities for the BOD photoswitch are very large compared to other molecular solar thermal storage systems and that these systems could be candidates for such applications.