People have been doing that for many years with the common hop vine (Humulus lupulus) which is also a member of the Cannabaceae family. Grafting hop vines onto a good Cannabis rootstock yields a scion with strobili that are visually indistinguishable from an ordinary hop flower. Unfortunately, the product is not very potent-- the best outcome is maybe 1.5-2% THC (and only trace amounts of other interesting compounds) which is terrible compared to the 10-20% THC that you can get from a well-managed C. sativa or C. indica flower. Also, the graft process is very finicky, the scion does not grow as well as an ungrafted vine, and your resulting plant is annual (like Cannabis) rather than perennial (like Humulus.) The hops you get are not terribly useful for beer-making, which is pretty much the only use for hops. (Some people like to make a sedative tea from hops, though I doubt that would be a good delivery method for the THC, since it's not water-soluble.) One other major "gotcha" is that the Cannabis plant matures much faster than hops, and the production density is hundreds of times better for Cannabis than Humulus.
Interestingly, there is some published scientific literature (see Crombie) that claims this grafting process does not work. However, I wonder, because Crombie talks about the hops "leaves" even though the only useful part of the plant is the flower (or properly, the "strobile.") The research I mention above has not been published, though the "1.5-2% THC" value I quoted has been measured by GC-MS. And, of course, there are just tons of anecdotal evidence from amateur gardeners that support either opinion.
I'll let someone else do the genetic research, but I think it may eventually be possible to engineer an algae that eats sunlight and poops THC. Wouldn't that be fun!