It looks like the authors of [3] misunderstood Chaitin. What Chaitin said about the limits of provability is that no statements of the form "K(x) > c_F" can be proven in formal system F where c_F is some constant depending on F.