s_fid is the fall back solution when s_vi can not be set. The situaion that is most common is due to locking occurring.
s_fid and s_vi, while not exactly interchangeable are still used as the u Kaur identifier for the row of data. So moving forward when you implement s_vi or ECID, etc it will still count the rows as unique rows and data would not have been lost. The new tracker will take the place of s_fid for users where it is currently set.
ECID or s_vi will take precedent over s_fid and no inflation would be seen.