If the single-world interpretation is not true, wouldn't it mean that quantum physics is useless?
What use is a quantum computer if the algorithm shows different results to different users?
Quantum computers are intrinsically stochastic machines. Quantum algorithms are about hedging the bets so that you get the "correct" answer at some p>0.5. Then you just repeat the computation until you're reasonably sure you have the right answer. In the Copenhagen interpretation the wavefunction magically "collapses" to one outcome, in Everett interpretation the wavefunction decoheres producing both outcomes in separate "worlds".
And what the use of a noisy channel if it may changes bits sent over it? ;)
You just need to have proper error correction. You are right - of course - if QM is completely noisy, or you can't have proper error correction (due to some physical limitation for example).