When we talk about ultramafic/mafic rocks we're really talking about ophiolites - sections of the seafloor which were produced at spreading centers and then shoved up onto the continental margin (instead of being subducted like they're supposed to). This is actually a pretty rare occurrence, and would lead us to believe that mafic and ultramafic rocks should be fairly uncommon. The Oregon Department of Geology and Mineral Industries Bulletin 95 (North American Ophiolites) has an Index Map (page v) which summarizes the locations of ophiolites in North America. There aren't many, and I estimate that somewhat less than 5% of the continental United States is underlain by them.

A special note concerning ultramafic materials is in order. This stuff is presumed to originate in the upper mantle, and represent what is left behind after the differentiation process separated out basaltic magma at spreading centers. This stuff (the ultramafics) is relatively heavy, really doesn't want to be anywhere near the upper crust, and must require some pretty intense (and unusual) forces to get it exposed at the surface. It is clear that most of what we lump together as ophiolites is actually the "mafic" portion (basalt, diabasic dikes, and gabbro), with a much smaller percentage consisting of the upper mantle ultramafic units. My gut feeling tells me that well less than 1% of the U.S. is underlain by ultramafic material (mostly peridotite, dunite, pyroxenite, and serpentine). Furthermore, I wouldn't be surprised to find that this approximate percentage holds at the global level. Strategic minerals (nickel, chromium, cobalt, and others) occur associated with these rocks, and their sparse and irregular global distribution indicates that ultramafic terrains are few and far between.