I haven't read any studies specifically on masking noise, but I've read in other studies that masking noise does decrease dysfluency for most stutterers. The theoretical explanation is that the error-monitoring systems in our speech production report too many false positives when listening to our own auditory feedback; so lowering the amount of auditory feedback can lead to less blocks. Here are some studies on auditory masking for anyone who is curious: https://scholar.google.com/scholar?hl=en&as\_sdt=0%2C22&q=stutter+masking+noise&btnG=