As Ed pointed out, there have been some recent grumblings in the ECMWF, GFS and CMC global numerical forecast models. HPC has picked up on this and felt it warranted a mention in their middle-extended range forecast discussion:

TROPICS
GFS SERIES HAVE BEEN INDICATING WRN CARRIBEAN TROPICAL DEVELOPMENT
FOR OVER A WEEK AT LONGER RANGES INDICATING INCREASING FAVORABLE
CONDS FOR TROPICAL DEVELOPMENT. LATEST OVERNIGHT RUNS AND TODAYS
12Z RUN ALONG WITH A LARGE NUMBER OF ENSEMBLE MEMBERS WITH WELL
ABOVE AVERAGE LOCATIONAL CLUSTERING CONT THIS TREND. 00Z ECMWF
ALSO DEVELOPS LOW PRESSURE IN THIS AREA BUT NOT TO THE DEVELOPED
EXTENT OF THE GFS. CMC ALSO LIKES THIS GENERAL REGION BUT PLACES
ITS EMPHASIS ON THE PAC SIDE. AFTN FINAL PROGS WILL DEPICT A GFS
ENS MEAN AREA OF LOW PRESSURE OFF THE NICARAGUA COAST DAYS 5-7
THU-SAT.
RAUSCH/ROSENSTEIN

I have found that the experimental Roundy cyclone probabilities product found here can be quite useful in determining at least favorable regions for development at extended leads. Currently, the product indicates that unfortunately for Berma, that region of the west and northwest Indian coast is dramatically coming into a favorable regime. The product also indicates that regions of the eastern Pacific are in a burgeoning region of potential development, too. Although this is not extended into the Caribbean, it is however close enough to the region, that combined with these operational global model signals, raises an eyebrow or two - particularly considering that nearing and going into June is also climatologically favored in this area and up throughout the Gulf.

...So, this is something to at least keep an eye on in future model runs.

John