<p>Tsunami waves are long gravity waves that propagate nearly conservatively in deep water, but may undergo significant energy dissipation during nearshore propagation due to coastal vegetation, roughness, and man-made structures. Quantifying this obstacle-induced attenuation in a physically consistent and transferable way remains a key challenge for the assessment of tsunami hazard. This study presents a comparative theoretical and experimental framework for the attenuation of tsunami-like long-wave pulses in vegetated or obstructed environments. Energy-based formulations are developed for both solitary waves and tsunami-like N-waves, demonstrating that they share the same hyperbolic amplitude-decay structure, with differences arising solely from waveform geometry through a shape-dependent coefficient. Within a common shallow-water setting, the analysis explicitly separates waveform representation from the adopted damping closure. These formulations are contrasted with linearized depth-averaged resistance models commonly used for long waves, considering both a constant-rate exponential attenuation and a pulse-consistent variant. Laboratory experiments involving solitary waves propagating through rigid stem arrays are used to calibrate bulk drag coefficients and provide a physically grounded benchmark. The calibrated resistance parameters are then applied, without re-fitting and without introducing new experimental data for tsunami waves, to predict the attenuation of tsunami-like pulses using alternative waveform representations and resistance closures. The results show that, for identical obstacle properties, predicted wave damping depends primarily on the adopted attenuation closure. In particular, constant-rate exponential formulations systematically overestimate attenuation for finite pulses, highlighting the importance of energy-consistent, pulse-based models for tsunami propagation in vegetated regions.</p>