To start, I will say that I am not an expert on FKR. That being said, I disagree with this on two points.
First, I don't think it's accurate to say that FKR has "hidden" rules. That implies some level of misdirection or deception. IMNSHO, FKR is best described as being based on trust: trust between the players, trust in the referee, trust in the game. The idea that there are parts of the game that are "hidden" seems antithetical to that premise. That a referee may need to come up with an ad hoc ruling to an unexpected scenario is not a "fail" state, it is entire reason for the referee to exist in the first place. Additionally, I don't think that two FKR referees could be expected to make two identical rulings, because the defualt assumption is that two completely identical scenarios should not exist.
Second, FKR is not inherently "rules light". The baseline assumption is that the players agree to the priorities of the game (i.e. world, narrative, etc) as well as the authority of the referee. The players need to know and agree to the rules, and trust the other players and referee to know them as well. In a majority of FKR games, the rules are kept intentionally light and simple to promote the other priorities of the game. However, there is no inherent need for the rules to be light, only that they are known and agreed to. You could, theoretically, play FKR with D&D 3e, the Battletech wargame rules, or the Arkham Horror board game. As long as the players understood those rules and agreed to them. Now, this is not commonly done. Largely because those rules don't translate to the other priorities of the game. Partially because it tacitly assumes a level of rules mastery that most players won't have. And, most generally, because people playing FKR tend to not want higher complexity rules. But none of that means that "rules heavy" FKR doesn't exist, just that it's not common.
To sum all that up: FKR has an extremely high correlation with rules-light play, but there is no causation that requires it.