From the beginning of the project, extensive numerical work was carried out using the Portuguese HPC cluster Deucalion (project 2025.00196.CPCA.A3) with the PIC code OSIRIS, enabling a detailed identification of the dominant mechanisms involved in the development of filamentation instabilities in laser-solid interactions. These simulations revealed that plasma collisionality, strongly temperature‑dependent, plays a crucial role in shaping the instability and leads to rich spatiotemporal dynamics, with plasma filaments primarily developing at the front surface of the target. They also highlighted the unexpectedly strong importance of electrostatic effects (ion motion), contrasting with the commonly adopted picture of hot-electron‑driven filamentation. These findings led to two major research directions: (i) a collaboration with IFN‑GV (UPM, Madrid) to model collisionality—absent in standard PIC codes—and to design new diagnostics capable of tracking its spatiotemporal evolution; and (ii) the development, both numerically and analytically, of a novel quasi‑static spatiotemporal framework for studying filamentation instabilities, culminating in a new analytical model capturing the full unstable electromagnetic spectrum. On the experimental side, a proposal informed by this modelling effort and the new collision diagnostics was submitted to the HED instrument at XFEL. Although not awarded beamtime in this first round due to limited availability, the exceptionally positive feedback strongly encouraged resubmission.