Phase-field fracture models are increasingly used for simulating brittle failure in laminated glass structures subjected to extreme loading conditions such as blast events. However, standard phase-field formulations rely on a single fracture energy parameter G_f, which simultaneously governs crack initiation and crack propagation. This intrinsic coupling prevents independent calibration of tensile strength, fracture toughness, and the regularization length scale, leading to a fundamental limitation in engineering applications. Under high-rate blast loading, the choice of G_f either enforces a realistic onset of fracture at the expense of excessive crack resistance, or yields physically admissible crack propagation while triggering premature damage initiation. To address this shortcoming, this contribution proposes a fracture-energy switching strategy within the phase-field framework. Two distinct fracture energies are employed: an initiation-controlled G_init governing damage nucleation, and a propagation-controlled G_prop activated after crack initiation. The transition between both regimes is driven by internal damage measures, enabling a partial decoupling of crack onset and crack evolution while retaining the variational structure of the model. In contrast to impact loading scenarios, where phase-field fracture formulations face additional challenges due to evolving contact conditions, blast loading represents a comparatively soft loading regime. The present study therefore focuses on blast-induced fracture as a representative high-rate loading case in which the intrinsic limitations of single-parameter phase-field formulations can be examined without the confounding effects of impact-induced contact instabilities. A systematic parametric study is conducted for laminated glass specimens subjected to blast loading, comparing full phase-field formulations with hybrid approaches and different decomposition strategies. The influence of the proposed switching mechanism on damage localization, crack patterns, and global structural response is assessed. The results identify modeling configurations that remain numerically stable and physically meaningful, and that are capable of surviving typical engineering calibration requirements. The proposed framework improves the applicability of phase-field fracture models in high-rate brittle failure scenarios, where conventional formulations remain numerically functional but lack sufficient predictive capability.