Graph theory techniques have emerged as useful tools in analysis of the physical performance of power grids, especially cascade failure events. Based on such models, several approaches have been proposed to estimate transmission network vulnerability. The most accepted methodologies rely on random error and deliberate attack tolerance assessment, conducted upon complex network paradigms. However, a lack of validation studies have been conducted for verification of the effectiveness of the results obtained with such methodologies, when applied to electric infrastructures. This paper investigates structural vulnerability by means of comparisons between physical power flow models and scale-free graph statistic indexes, generating conclusions on the appropriateness of graph theory techniques for assessment of electric network vulnerability. This is performed through simulations on standard IEEE bus-testing networks. A discussion of the utilisation of several indexes in scale-free graphs for vulnerability assessment is also contributed.