The rapid growth of electric vehicles (EVs) and stationary energy storage is accelerating the deployment of lithium-ion batteries (LIBs), including their operation in thermally unmanaged environments. Computational engineering methods can support scalable assessment and control through reduced-order models that deliver physics-informed features at low computational cost. A key limitation is the lack of unified comparisons of temperature-dependent, reversible charging losses across widely used commercial LIB chemistries under identical protocols. This study addresses this gap by combining controlled thermal experiments with reduced-order impedance modeling. Commercial 18650 LIB cells are selected based on the widespread adoption of their cathode chemistries: lithium nickel cobalt aluminum oxide (NCA), lithium nickel manganese cobalt oxide (NMC), and lithium iron phosphate (LFP). Cells are charged using constant-current/constant-voltage (CCCV) protocols at five ambient temperature setpoints ranging from 10 to 45 °C. Discharge is maintained at 25 °C to isolate charging effects. Reversible capacity is assessed together with direct-current internal resistance (DCIR), alternating-current internal resistance (ACIR), and electrochemical impedance spectroscopy (EIS) over 0.1 Hz to 10 kHz. The impedance response is represented using a reduced-order 2RC equivalent-circuit model to extract compact descriptors of ohmic, interfacial, and diffusion-related processes. Results show that low-temperature charging causes the strongest reversible capacity loss in NCA cells, a moderate loss in LFP cells, and the smallest loss in NMC cells. Charging performance improves at higher ambient temperatures for all chemistries. Internal resistance decreases as temperature increases and follows the same qualitative trend as charge acceptance. Reduced-order 2RC parameters capture chemistry-dependent polarization pathways and provide computationally efficient features for surrogate modeling, comparative screening, and temperature-aware charging control in EV and grid applications.