No clear example has been produced where one can imagine that p and there is also a good argument that it is impossible that p

Before Kripke (1972/80), the first challenge to such an argument would have concerned the move from (3) to (4). When philosophers generally believed in contingent identity, that move seemed to them invalid. But nowadays that inference is generally accepted and the issue concerns the relation between imaginability and possibility. No-one would nowadays identify the two (except, perhaps, for certain quasi-realists and anti-realists), but the view that imaginability is a solid test for possibility