The introduction of the paper mentions some alternative interpretations put forward in earlier papers. It seems as if this "problem" has been noted earlier. This new solution to the problem beats earlier solutions. But maybe it is an evolving discovery process which again will invent new better explanations?
In addition to those mentioned in the paper, I suggest the following candidate alternatives:
-) Too few observations in order to hold up in a soon dramatically increased discovery rate. 4 or 5 or maybe even 8 TNOs which, as they claim, give a 0.007% significance level is certainly resting on some assumptions which will be challenged if thousands of similar objects are found in the next decade. So, as Andy has answered: possibly a coincidence due to small sample. (The paper's 0.007% probability is, well, we'll see)
-) Weird unthought-of science bias which, in math more than in lens, tends to select inadvertently among the (candidate) observations made and their characteristics. Since the authors are at the top in their fields, and Michael Brown has discovered loads of distant objects, including Sedna, because of that maybe one could suspect that some one thought has created a bias somehow.