In this case, poor Holo has been transected by an evil surgeon. Both Holo-Left and Holo-Right contain, for the most part, the memories, personalities, and other mental traits of the original Holo. The interpretation, then, is straightforward: both Holo-Left and Holo-Right are (mostly) the same person as Holo.
Bruce makes much of the division of the brainstem, since it supports the sleep/wake cycle. By supposing that half a brainstem cannot perform this function, he concludes that the evil surgeon must have "cloned" another brainstem half and (by this cloning) created a new person. While the supposition is reasonable from a scientific standpoint, philosophically it merely dodges the question. I suggest that the principle of numerical identity, which insists on counting persons as indivisible, uncopyable units, cannot adequately deal with this case. In cases involving copying, one can assert that the result is simply a new person; but in this case, there is no copying, and so this argument cannot apply.
As time goes on, what is the relationship between Holo-Left and Holo-Right? Clearly, they have different experiences, which will cause them to become progressively different people. They will be mostly the same person for quite some time, however, and always more the same person than, say, twins can ever be, due to the time they shared when they were Holo. Boolean logic cannot handle a case like this, where the participants are both partly the same person and partly different people.
Patients whose corpus callosum has been severed present an interesting (though somewhat muddy) real-world analogue of this case. Such patients are usually able to function normally, though as Bruce correctly points out, there are sometimes conflicts between the two hemispheres. In some cases, there appear to be two separate persons sharing the body, with somewhat different memories, goals, and skills. Any philosophy (e.g., causal/numerical identity) which is constrained by Boolean logic must struggle to force these patients into categories designed in the context of "normal" whole-brain people. To describe them accurately, however, we must be more flexible and recognize that the two hemispheres give rise to entities which are mostly, but not entirely, the same person. [Split-brain patients are considerably better off than Holo-Left and Holo-Right for two reasons: they have united lower brain structures and other interhemispheric commisures, and except in carefully controlled conditions, both halves tend to experience the same events from nearly the same viewpoint.]
The case of Holo-Left and Holo-Right is a very instructional one. It demonstrates clearly the problems with traditional views of identity, and the necessity for a more flexible, "fuzzy" approach.