What organism/assembly is it? If it's a UCSC genome, I can wrap the package for you.
Otherwise put the package source tarball somewhere where I can access it. This is how we've managed contributed BSgenome data packages so far.
Note that the plan is to eventually refactor the current approach and have BSgenome data packages pull the sequence data from AnnotationHub instead of having it embedded in the package. This has been on the radar for a while, just need to find the time to work on this.