Single molecules (SMs) have become important tools for studying nanoscale dynamics in biological studies since they were first optically observed in 1989. Numerous techniques, termed single-molecule orientation-localization microscopy (SMOLM), have been developed to measure the position and orientation of SMs. Due to the challenging signal-to-background ratio (SBR) in typical SM experiments, it is critical to choose/design an imaging system that is optimal for a specific target sample. Inspired by the performance limits of SMOLM techniques we derived, we develop new methods for measuring the orientation and position of SMs. First, we report a radially and azimuthally polarized (raPol) microscope with high detection and estimation performance. Imaging Nile red (NR) molecules transiently bound within DPPC supported lipid bilayers (SLBs) reveals the existence of binding pockets that limit NR from freely exploring all orientations. Treating the SLBs with cholesterol-loaded methyl-β-cyclodextrin causes NR’s orientational diffusion to become significantly more confined. Strikingly, NR's translational diffusion drastically increases despite the cholesterol-induced condensation of the SLBs. We also developed a multi-view reflector (MVR) microscope for 3D SMOLM. The localization and orientation precision using a radially and azimuthally polarized MVR (raMVR) microscope is ~2 times better compared to other state-of-the-art SMOLM techniques. Imaging lipid-coated silica spheres, we show that the raMVR microscope can resolve the 3D positions and orientations of NR molecules transiently binding to spheres as small as 150 nm and as large as 1 μm. Further, we experimentally show its robustness against aberrations caused by refractive index mismatch. The raMVR system also allows us to detect and discriminate between lipid-based membranes versus amyloid aggregates by measuring the rotational dynamics of NR molecules. These detailed measurements of SM rotational and translational dynamics are made possible by raPol and raMVR's high measurement performance, and we expect the raPol and raMVR systems to be adapted in many future SMOLM studies to reveal higher-dimensional dynamical processes in live cells.