Offset-driven self-retraction and oscillatory motion of bilayer graphene has been observed experimentally and is potentially relevant for nanoscale technological applications. In a previous article, we showed that friction between laterally offset graphene layers is controlled by roughness and proposed a simple reduced-order model based on density-functional theory (DFT) and molecular dynamics (MD) data, with which predictions on the experimental size-scale could be made. In this article, we extend our study to other layered materials, with emphasis on boron nitride (BN) and molybdenum disulfide (MoS2). Using MD and DFT simulations of these systems and a generalized version of the reduced-order model, we predict that BN will exhibit behavior similar to graphene (heavily-damped oscillation with a decay rate that increases with roughness) and that MoS2 shows no oscillatory behavior even in the absence of roughness. This is attributed to the higher energy barrier for sliding in MoS2 as well as the surface structure. Our generalized reduced-order model provides a guide to predicting and tuning experimental oscillation behavior using a few parameters that can be derived from simulation data.