The van der Waals gap of two-dimensional (2D) materials provides natural ion diffusion channels for lithium. However, individual 2D materials (such as graphene, ${\mathrm{MoS}}_{2}$, and MXene) cannot offer all the properties needed to maximize the performance for energy storage. Heterostructures, by stacking different 2D materials, are promising for combining the advantages and eliminating the disadvantages of individual 2D materials. ${\mathrm{MoS}}_{2}$-based heterostructures have been prepared for lithium storage, while intercalation and diffusion mechanisms have yet to be elucidated. In the current work, the heterointerface effects on lithium intercalation and diffusion of seven $H$- and distorted $T$ $(\mathit{dT})$-${\mathrm{MoS}}_{2}$-based heterostructures are systematically studied by density functional theory calculations, and highlight the atomistic origin of differences between $H$$(\mathit{dT})$-${\mathrm{MoS}}_{2}$/graphene, $H$$(\mathit{dT})$-${\mathrm{MoS}}_{2}/H$$(\mathit{dT})$-${\mathrm{MoS}}_{2}$, and $H$$(\mathit{dT})$-${\mathrm{MoS}}_{2}$/MXene hetero- and homostructures. From the energetic point of view, the heterointerface makes for an easier phase transition of ${\mathrm{MoS}}_{2}$ to the $\mathit{dT}$ phase in the ${\mathrm{MoS}}_{2}$/graphene heterostructure. Then we discuss the general idea of a rational design of the van der Waals gap for efficient lithium storage and diffusion. Our work opens a new avenue for optimizing the ion diffusion properties of 2D material heterostructures by designing the van der Waals channel for ion diffusion.