hi Gabriel,
using the Fit Model with the specifications you mentioned is what i did in the past to run the model and it did run with my data using JMP 9 at the time. another thing i use is the overdispersion tests and intervals and the Firth bias-adjusted estimates. i actually run the models twice with and without them to inspect robustness.
From my experience, if something in the Fit Model platform is not working properly, i try fitting each X separately to the Y to determine which of the variables is the source of the problem.
ron