I can only attribute it to a progressively degrading connection at that time. The bit that surprises me is that there wasn't any device attached, just the cable.
this is why it is, and most likely remain, a mystery. Problem is if it was contact resistance/metal migration on point contact.. it is a series resistance.. is only of significance when a device is attached.. but when no load is applied the open circuit resistance infinite (well at least the insulation resistance in the usb lead....

This is obviously not a h/w issue...if it was why will windows boot into safe mode?
Well yes it still can be H/w related.. windows may not know about the H/w issu untill it attempts to access the port..or turn the h/w channels on.. retrieve data from the device... Not all faults are SHORT CIRCUITS.. Nor are all faults open circuits..
But in safe mode many of the hardware drivers are not functional..that is why it is called safe mode.... except I thought USB drivers are enabled in safemode in win XP and 2K... curious still