Anyways, I doubt that plant nuclear genome size has any relevance, since you are interested in the organellar genome, and those are usually never longer than 1 Mb, usually couple hundred kb. Although the smrtcell would certainly be useful to close gaps, you may need some additional sequencing with another technology perhaps to get a high quality consensus sequence. The question to answer is what is the % of reads that will be chloroplast and not nuclear, I suggest looking at other studies.

I told you why I closed it, it has nothing to do with bioinformatics and everything to do with sequencing strategy and library construction. Hence I closed it. As it happens you got an informative answer from someone at PacBio on SeqAnswers: http://seqanswers.com/forums/showthread.php?t=65357