I'm not sure we can correctly do the statistics atm considering we don't know how often it's intended for the variance to swing to the positive or negative modifier or if their intent is that the 5% is a flat or maximum cap value. Regardless you're right in the sense we'd probably get something like Hunter One doing 97,500 dmg (lower variance) over a fight whilst Hunter Two is closer to 100,000 (average) so by value alone 2500 dmg less doesn't seem like much but my point was the logs would show different. As in someone is going to look at the logs see that Hunter One performed worse than Hunter Two and assume player error over the system screwing their damage rolls. Especially so if there are more Hunters in the raid that end up in variance between or even higher than the two hunters and this is without even considering how crits, procs and such would factor into it.
Well the hypothetically is assuming static damage without taking into account procs and how they might be changed going into Shadowlands which is why I wasn't basing it on the current fluctuating damage numbers. Even then if they continue to stay the same the variance system is still likely to mess with damage rolls especially in regards to crit values.
So assuming an average of 1000 damage you don't think it's likely you could low roll and get 950, 980, 960, 955, 971 etc consistently throughout a fight? Have you played a Blizzard game?