Despite the rapid expansion of Mobile Laser Scanning in forestry, the optimal deployment of affordable handheld LiDAR systems for forest characterisation, particularly at understorey level, remains underexplored. The structural complexity of Mediterranean ecosystems requires highly detailed spatial data to estimate fuel loads, biomass and ecological functioning. This study addresses this gap through an experimental comparison of the open-source MandEye LiDAR device against a premium commercial system, the FARO Orbis Premium, functioning as the ground truth baseline. The research evaluates how pedestrian scanning trajectories, including single loops, multiple concentric loops and complex patterns, affect point cloud quality, understorey volume estimation and structural accuracy within a 10 × 10 m Mediterranean pine forest plot. Data processing involved point cloud normalisation, tree stem excision to obtain the understorey layer and volumetric modelling via voxelisation and alpha–hull algorithms. Results indicate that trajectory geometry fundamentally dictates SLAM algorithm performance and data reliability. A double concentric loop proved optimal for both devices, overcoming structural occlusion while minimising redundant noise. Conversely, complex trajectories severely penalised SLAM stability, inducing positional drift and structural duplication that wrongly modified and inflated volume metrics. Although the sensor in MandEye exhibited lower canopy penetration compared to FARO, which resulted in a systematic underestimation of internal understorey volume (e.g., 3,31 m³ bias underestimation during optimal double-loop voxelisation), it successfully captured highly dense and continuous external canopy representations when operated along predictable paths. Ultimately, this study establishes critical methodological guidelines for deploying accessible mobile scanners in forestry.