Errors come in groups on different places. I think about "timeframe" view, where you would specify few seconds window in time and you will get all errors from various sources (log files, log.nsf, openlog...) for that period of time.
Offsets in times in different logs could be a problem (regional setting, DLS, server/client time)...