The lone-pair…π (lp…π) (deoxy)ribose…nucleobase stacking is a recurring structural motif in Z DNA and RNAs that is characterized by sub-van der Waals lp…π contacts (<3.0 Å). It is part of the structural signature of the CpG Z-steps in Z-DNA and r(UNCG) tetraloops. These nucleic acid structures are poorly behaving in molecular dynamics (MD) simulations. Although the exact origin of these issues remains unclear, a significant part of the problem might be due to an imbalanced description of non-bonded interactions including the characteristic lp…π stacking. To gain insights into the links between lp…π stacking and MD issues, we present an in-depth comparison between accurate large-basis-set double-hybrid Kohn-Sham density functional theory calculations DSD-BLYP-D3/ma-def2-QZVPP (DHDF-D3) and data obtained with the non-bonded potential of the AMBER force field (AFF) for NpN Z-steps (N = G, A, C, U). Among other differences, we found that the AFF overestimates the DHDF D3 lp…π distances by ~0.1-0.2 Å while the deviation between the DHDF-D3 and AFF descriptions sharply increases in the short-range region of the interaction. Based on atom-in-molecule (AIM) polarizabilities and SAPT analysis, we inferred that the DHDF-D3 vs. AFF differences partly originate in the Lennard-Jones (LJ) parameters that are identical for nucleobase carbon atoms despite the presence/absence of connected electron withdrawing groups that lead to different effective volumes or vdW radii. Thus, to precisely model the very short CpG lp…π contact distances, we recommend revision of the nucleobase atom LJ parameters. Additionally, we suggest that the large discrepancy between DHDF-D3 and AFF short-range repulsive part of the interaction energy potential may significantly contribute to the poor performances of MD simulations of nucleic acid systems containing Z-steps. Understanding where, and if possible why, the point-charge-type effective potentials reach their limits is vital for developing next-generation FFs and for addressing specific issues in contemporary MD simulations.