I have a question essentially the same with this one "Aharonov-Bohm Effect and Flux Quantization in superconductors" which is why we can say the flux is quantized in superconducting disk but not in AB effect.
I think David Bar Moshe's answer is really reasonable: In AB effect the wave function is the real wave function of the electron while in superconductor case the pair potential $\Delta(r)$ is actually a correlation function thus it must be single valued.
However, this argument seems cannot explain the Laughin's argument of IQHE. Since in his argument it is the wave function of the extended state(which is the real state of electron not macro correlation) around the loop forces the flux to be quantized.
Does anyone have any better explanation when we can have quantized flux and when cannot?