Well, the breakup of Pangaea only started for biogeographical purposes around the Early-Late Cretaceous boundary when Africa broke off of South America and apparently lost the spinosaur connection to Europe (...and Thailand).
Not sure what you mean here. The Laurasia-Gondwana split occurred a long time before this (Late Jurassic).
Not for biogeographic purposes. One word: *Baryonyx*.
As for dromaeosaurids, their distribution was compatible with a vicariance model, so long as the Cretaceous unenlagiines were limited to Gondwana and the velociraptorines and dromaeosaurines were limited to Laurasia. But when Turner et al. (2007; the _Mahakala_ paper) recovered the Mongolian taxon _Shanag_ within the Unenlagiinae, this throws a spanner in the works, because it implies at least one dispersal event.