Children do not understand concept of false beliefs until age 6 or 7

I was a bit suspicious, so I had to come back to this after sharing the "updated" 3-box question with my not-yet-5 year old. I've paid attention not to use the wording like "look first" to avoid any suggestive approaches (which is highlighted as giving much better outcomes in the study).

Prompted about why he chose the right box where a kid will look for their chocolate, he immediately blurted out "because that's where she left it" as if I am asking a foolish question. To me, that removes guessing from the equation altogether which is what study claims children at this age (before 6-7) will do 100% of the time ("proven" by it being an exactly 50%-50% split between two empty boxes).

To be honest, I am not sure why was I intrigued by this study in the first place. I suspected different outcomes come from child's exposure to logic fallacies (I frequently joke with my kid by saying obviously wrong things and let him challenge me, so he is used to them), language and time concepts (I find my kid frequently confusing "yesterday" and "tomorrow", for instance, and these questions had a lot of "relative time" descriptors — I can imagine many other kids have the similar issues with what happened first and what happened after), confidence (my kid, being a shy one, might not be as assertive with researchers), hastiness (my kid will sometimes guess at things just to get me off his back when I am asking something if he doesn't care about it) etc.