When looking at curves like these or for that matter all-in adjusted EV, its important to remember, that they only measure a selected part of variance or luck. Our results will also be influenzed by, how often our opponents make big hands, when we are playing with them, and by the distribution of coolers like AA vs. KK or set over set. Or even just how often we hit top pair vs. miss completely.
Also a standard deviation will be way smaller after 200k hands than 10k hands. So even if the graph show, we have been running cold with sets, flushes and straights over 200k hands, that might only mean, we have hit them a few percent less often than statistically expected. Say we only saw a flop with a pocket pair 8 times. Then we were expected to flop a set once, but of course many samples will see us flopping a set 0 or 2 times and occationally 3+. So even a large relative deviation like seeing 100% fewer or more sets wont put us far out on the bell curve, because its a fairly common outcome.
If however we saw a flop with a pocket pair 10.000 times, then flopping a set 1.100 times or 1.400 times rather than the expected 1.250 times will already put us far out on the bell curve, because not many people will run this cold or hot. But its not like, the unlucky player never saw a set. He just did not see quite as many of them as statistically expected. So if we have bell curves like those of GreenDaddy1 for 6-max games after 200k hands, thats obviously not great. But at the same time it probably also did not have that huge of an impact on our overall results. If we are beating the games, we should be able to show a profit even with this little bit of headwind.