Human activities have increased species extinction rates as much as 100-1000 times the historical background rates over the past centuries. Understanding how anthropogenic activities impact future biodiversity is of vital importance. The models we use for this purpose often exhibit a trade-off between taxonomic coverage and data requirements. At local scales, detailed models can be made for few well-studied species for which demographic parameters are available. In contrast, large-scale models increase generality (geographic and taxonomic coverage) at the expense of ecological realism. This study aims at optimizing species-based models at both ends of this trade-off by increasing generality or ecological realism.