Something I experienced a few times.
I had a old Bigfoot hard drive that decided to die.. sloooowly.
It would get i/o input output errors
Read/write errors
This was pretty much the days before S.M.A.R.T. logging and utilities.
Of course, I had the benefit of running linux of the time, and I saw every time the OS or myself would try to access something on that hard drive the errors would come up. Depending on how big the sector is, and how large the section is that has the errors, it can slow you down significantly, or almost not at all. In my case, the moment I logged into Linux, I'd get these errors endlessly, and if I was lucky, 30 minutes in It would clear up.
How this translates to Microsoft operating systems is simple. The operating system simply keeps trying until it fails. Remember the old 'retry/abort/fail' messages in dos? Windows simply defaults to 'retry' until it decides, in my experience to 'fail' or 'abort' the process. Unfortunately you get nothing visually, except long wait and lag in data access.
Seeing as the computer, the operating system lands data at random for the most part on the hard drive, its hard to say, without a chkdsk, or scandisk to lock out those damaged areas (or more recent HD tools).
If you manage to move the data, defrag it, to another location, you might run 5 years without problems, or 1 hour, or 1 minute, depending on the OS. Figure on replacing the hard drive, now, and move the data over that you want to save.
Good Luck