What is the best (reference) quantum chemical approach to decipher the energy components of the total interaction energy : Symmetry-Adapted Perturbation Theory (SAPT) or Supermolecular Energy Decomposition Analysis (EDA)? This is a very common question that force fields developers ask themselves when designing new intermolecular potentials. With the rise of physically motivated polarizable force fields able to include various type of physical effects within molecular dynamics simulations, the need to answer such an interrogation becomes critical. In this paper, we perform a systematic and detailed assessment of three variants of SAPT (SAPT2, SAPT2+3 1 arXiv:2008.01436v1 [physics.chem-ph] 4 Aug 2020 and SAPT(DFT)) and three supermolecular EDAs approaches (ALMO, CSOV and RVS) on a set of challenging, strongly bounded water complexes with cations and anions. Using a regularization scheme within SAPT(DFT), denoted Reg-SAPT(DFT), to partition the second-order induction energy into its polarization and charge delocalization contributions, we show how the single-exchange, or S 2 , approximation has a large effect on these energies, and provide additional evidence for not using the initial Stone-Misquitta definition for charge delocalization. Alternatively, we obtain more satisfactory results for the evaluation of the polarization and charge-delocalization energies using an infinite order strategy. As we compute these quantities using SAPT(DFT), Reg-SAPT(DFT) and classical polarization models, we show a convergence with supermolecular EDAs, despite the observation of sizable residual differences between methods. Overall, for strongly polar systems, when separable physically motivated energies are needed, we recommend the use of SAPT(DFT) without the S 2 approximation, the supermolecular EDA ωB97X-D||ALMO being our second choice. As neither SAPT2 nor SAPT2+3 can be fully freed from the S 2 approximation, we do not recommend either of these methods for a separable, physically motivated decomposition of the interaction energy. The results of this paper propose some practical recommendations for SAPT(DFT) calculations and are organized in the form of a check-list aimed to be useful for force fields developers.