Both points are invalid in this case.
The core audience of this game spent so many hours in different looters they usually develop the feel for when RNG is skewed. I can’t say for OP or any other player’s experience or issues, but my experience - in games and RNG in general - and my education allow me to say GD has RNG anomalies outside of observer’s error margin.
Sample size indeed matters, you can’t say “this item doesn’t drop in this game” based on your playtime. But for detecting anomalies sample size has little meaning. For instance, if you enter the dungeon and of all the blue items in the drop pool you have one single item dropped 4 times in a row you don’t need 10.000 more runs in this dungeon to say smth is wrong. Because you already have a statistically impossible example.