Comment by antognini

I've heard this claim, but I don't think it holds up. The idea is that supposedly the calculation is done in parallel across the different universes which allows you to see an exponential speedup in some calculations. But the whole point of many-worlds is that after a branch event the various branched portions of the wavefunction no longer interact. Even if it were true that quantum computing was really a bunch of parallel classical computations (which it is not), there would be no way for any of the other universes to communicate the result back to the universe you observe it in.