Systemlog entries are always in chronological order so even if you notice a timestamp issue, you know that it was issued between its surrounding correctly timed commands.
If the large descrepency bothers you then on a scheduled reboot this could close the gap significantly
1)to save the time into a file
2)soft reboot
3) zero the time with the time in the file instead of instead of epoch time
4) delete the file immediately after so the time doesn't backdate on any unexpected power cycle
(It will still be offset for however long the reboot takes to complete until next ntp update)
--
OR
Check system log after bootup for ntp descrepency notifications.
Rewrite system log updating all epoch times according to discrepancy found so they are all correct.