Are there any non-interpretational arguments in favour of Everett's Many-Worlds? As far as I understand Everett's Many-Worlds Interpretation, he makes the case for a realist theory of QM at the enormous cost of many-worlds.
Are there any arguments in favour of Everett's interpretation that makes the case for his theory one a purely instrumental or operational view - in that certain calculation look easier or more 'elegant'?
 A: I don't really know what the questioner means by whether the MWI makes a calculation easier or more elegant. I can think of calculations that cannot be done in any other interpretation because they deny that quantum mechanics applies to macroscopic  objects. For example, David Deutsch and Patrick Hayden explained how correlations are established between the results of measurements on entangled quantum systems. The correlation is not established until the measurement results are compared and this can happen because decoherent systems can carry quantum information provided that it doesn't affect the expectation values of measurements until the comparison is done, see:
http://arxiv.org/abs/quant-ph/9906007
http://arxiv.org/abs/1109.6223.
A: The Stanford Encyclopedia of Philosophy's article on Everett gives a much better description than Wikipedia. "Everett's solution to the [measurement] problem was to drop the collapse postulate from the standard formulation of quantum mechanics then deduce the empirical predictions of the standard collapse theory as the subjective experiences of observers who were themselves modeled as physical systems in the theory." Everett's formulation does not have to be viewed in terms of the pop-sci conception of many worlds. The article describes it as a no-collapse formulation.
At the end of Everett's thesis he writes "It remains a matter of intellectual interest that the statistical assertions of the usual interpretation [of QM] do not have the status of independent hypotheses, but are deducible (in the present sense) from pure wave mechanics, which results from their omission." His formulation takes more computational work to produce probability predictions since they are derived and not inherent in the theory. The point is its conceptual elegance doing away with wave collapse.
A real argument for Everett's interpretation would be a testable prediction. The encyclopedia of philosophy article claims that "Everett held that it was always in principle possible to measure an observable that would detect an alternative post-measurement branch." I have no idea if anyone has ever tried to make a concrete prediction from this formulation.
A: Definitely, there could be calculations, that may look differently in different interpretations. But this is still interpretational. Even if the calculation changes, the end result does not.
If in Copenhagen the calculation may include several stages with a collapse at each step, in MWI one can calculate the total wave function (including the observer) and find the probabilities of certain results from its projections. Thus the total number of steps may reduce.
A: David Deutsch in his article "Quantum mechanics near closed timelike lines" thinks that the MWI, and other similar interpretations, will give different results than Copenhagen, Bohm or what have you if you have closed timelike curves. So if you have a time machine, it is as simple as that!
