I'd call that a borderline case. All this stuff is speculation, so I'd speculate that a spliced together genome would probably be identifiable as such. If we know enough about the functioning of a genome to build a new one, we'd probably know enough to be able to recognize an artificial one.