Hello there!
I can't say that I have seen this issue before, at least when you are speaking of hundreds of duplicates. I was going through our cases and the developer backlog, and I do not see any report on this.
We could potentially improve our method of obtaining the DeviceID. Currently, we query both Win32_USBControllerDevice and Win32_USBController and take just the last part. For example, if the DeviceID is USB\VID_0BDA&PID_5532\200901010001, we take just the last bit (200901010001), which in some cases leads to false duplicates if the last bit matches the DeviceID of another device. So, as an improvement, we should use the full DeviceID.
However, hundreds of duplicates do seem to be a lot!