That's really annoying but unfortunately likely to fall within the error margins, I "know" that traditional finger prick meters have a accuracy of +/- 15%. Which means when you measure 10 it might actually be 8.5 or 11.5 or anything in between. Don't know about the dexcom but libre says to be most accurate in the "normal" range whatever that may be. As far as I know it's not really specified. Say both have the +/-!5% error margin it might just be a ~1 point difference if the libre is on the low side and the dexcom on the high side. Personally I have the libre and two traditional meters. Frankly I stopped comparing them because it drives you crazy. One time libre is the lowest and meter 1 in the middle and meter 2 the highest. An hour later it can be reversed... With cgm's most would probably say it's the trend not the reading you should be looking at.
Knowing the libre, the manual scan is a minute value at the time of scanning the. later and better, result on the begin screen (?) is a 15 minute average with should be more accurate. Again, don't know how the dexcom works. But given momentarily scans and error margins they might just be both "right" I'm afraid.