Abstract We present the second iteration of the caramel-gas code, an empirical model of the broad-line region (BLR) gas density field. Building on the initial development and testing of caramel-gas, we expand the meaning of the model parameter α, which initially represented only the power-law index of the dependency of emissivity on radial distance. In this work, we test a more generalized radial power-law index, α, that also includes a description of the effective emitting size(s) of the BLR structure as a function of radial distance. We select a sample of 10 active galactic nuclei (AGN) from three different Lick AGN Monitoring Project campaigns to further validate the caramel-gas code and test the generalized radial power-law index, α. Our results confirm that the caramel-gas results are in general agreement with the published results determined using the original caramel code, further demonstrating that our forward modeling method is robust. We find that a positive radial power-law index is generally favored and propose three possible scenarios: (i) the BLR structure has increasing effective emitting size(s) at larger radial distances from the central source, (ii) emission is concentrated at the outer edges of the BLR, and (iii) stronger theoretical assumptions are needed to break the degeneracies inherent to the interpretation of reverberation mapping data in terms of underlying gas properties.