I am certainly not well-versed in metaphysics, so please correct the terminology I have employed throughout.
When I speak of a total explanation of reality, I am describing an explanation that accounts for all—or, the total—of metaphysical reality. If there were anything over and above that explanation, than the original explanation would not be a total explanation at all. If a fish, for instance, accounted for the entirety of his bowl, that would clearly not suffice as a total explanation of reality. Why? Because there is a further explanation that is demanded by a non-local reality, that is, a reality that is not the fish's local reality, the outside world (of course, this assumes that the fish has no consideration or observation of the world outside his tank).
Now, I am wondering whether it is even epistemically tenable to postulate any kind of total explanation of reality. Any hypothesis we formulate is based on the things we can observe and know. Yet, there is nothing incoherent in the idea that there are many things that we cannot know that exist entirely disconnected from the universe in which we find ourselves. Thus, we cannot, it would seem, posit with justification—rather than pragmatism—these kinds of total explanations. Naturalism, for example, cannot be defended in the ultimate sense, because there may very well be things that are not natural that have not and cannot be observed beyond the confines of our universe. I imagine this also places many theistic religious views in a pickle as well. Christianity is defended as a total explanation, not a local explanation (it is not a case like the fishbowl). Thus, unless literally every metaphysical commitment a Christian theist makes is justified, Christianity cannot be rationally chosen over competing hypotheses that accommodate the incomplete set of commitments that have been justified. For example, taking the resurrection evidence as strong, theories like a kind of aesthetic deism where God has a laugh raising a random man from the dead or perhaps like a digitally simulated universe in which the resurrection is made to occur cannot ultimately be embraced or rejected since they accommodate the data just as well. Thus, Christianity would become untenable since there are ever so many ultimate hypotheses that have the same evidential pattern.
I suppose this goes into separate issues—namely skepticism and strong agnosticism—, but I am first wondering whether my inquiry as to whether a total explanation can ever be ultimately epistemically tenable is correct. Now, I do think one way out would be to prove that an omniscient being exist, and that would allow you to elude the grasp of locality; but otherwise, I do not see how these theses can be ultimately defended. Second, I am also wondering about this whole idea of selecting between competing theories that accommodate the data just as well as others. The idea, of course, can also be applied to issues like the problem of other minds. Thoughts?