MosSCI with large ( >16kb) genes?

I would like to know if anyone has any experience with using the MosSCI technique on large genes.

The FAQ says that it is possible to integrate inserts at least 16kb. The genes (with tags) that i am interested in are between 16 and 18.5 kb. Is this too large? Or will i just get very few lines?

Also, if it is possible to do this with my genes of interest, would I have to use the selection based insertion method?

any help is greatly appreciated,
jason