Technology Review has a post on a British effort to engineer the perfect potato. As far as I can tell from a quick read, it involves identifying potato varieties with the desired traits (blight resistance, etc.) and the genes involved, and combining them into one potato. Apparently there are "genetically modified" varieties already, each with a desired trait, so it's a logical next step to combine them.
When they write "genetically modified", I'm assuming it's not inserting genes from one species into another, but rather moving the genes in the laboratory, not by cross-breeding. It raises the question I've noted before: where do you draw the line in opposing GM-foods? At one end of a continuum is a plant/animal which is different than any which lived before, because the combination of genes is new, but one created by normal sex/seed production. Then you get into conventional breeding. Then moving genes in the lab, but still within the same species. Then using CRSPR to edit genes out. And finally adding genes across species lines.
IMO you can make the same cautionary argument in each case--there might be harm to humans from this new combination of genes. Obviously the likelihood grows as you move along the continuum. Again in my opinion I don't think there's much likelihood of harm at any point.