This isn't a philosophy of religion paper, but it's an absolutely essential read to those working in the field. Possible-worlds analyses (e.g. of concepts, properties, essences, dispositions, etc.) are now widely seen to be too coarse grained to "carve reality at her joints", as it were. Therefore, the future of analyses for God's attributes, of the viability of divine command theory, of the standards for what counts as a counterexample, etc., is likely to be a hyperintensional one, in which case consideration of impossible worlds will more frequently be seen as relevant to their evaluation. Let the hyperintensional revolution continue!