Recent advances in consumer mobile devices equipped with solid-state LiDAR have enabled rapid indoor 3D reconstruction for augmented and virtual reality (AR/VR) applications. This paper presents a systematic comparative evaluation of smartphone-based LiDAR scanning and photogrammetric reconstruction for indoor environments, benchmarked against terrestrial laser scanning (TLS) reference data. The study analyzes geometric accuracy, completeness, trajectory drift, visual fidelity, and real-time performance under AR/VR and 5G edge-computing constraints. Experimental results demonstrate that smartphone LiDAR achieves stable centimeter-level accuracy and real-time meshing suitable for immersive XR workflows, while photogrammetry provides superior surface detail and texture quality at the cost of higher processing latency and sensitivity to lighting conditions. The paper further identifies hybrid LiDAR–photogrammetry pipelines as a promising strategy for balancing geometric reliability and visual realism, supporting scalable digital-twin and educational XR applications.