Pedantic
Legend
That's really only true if that fortune is going to be evaluated as a direct comparison between the parties involved, or if both parties are making a comparison against a static obstacle. Attack vs. AC breaks down if the characters can't hit or be hit, and a wall that must be climbed, a rogue can definitely climb and a cleric definitely can't, for example.The truth though is all game systems are designed around situations where the 'bonus' to a roll is significantly smaller than span of fortune available from the fortune mechanic. In D20 this means the design is built around situations where your bonus on the D20 check is rather smaller than 19, and probably not larger than 10 or so. Any game system starts breaking down when the fortune is a rather small input to adjudication or where the range between bonuses of participants becomes large. It's not just a D&D thing. It has to do with the mechanics of playing well together and having reasonable expectations of failure or success.
Ideally the game design task should be using fortune in a more interesting way than those direct comparisons. It's not a problem if wall climbing is an ability only in one character's skill set, if the situation at hand has more points of interaction and could be solved in multiple ways (and especially if climbing is only one step in a broader set of necessary actions). The worst possible use of randomness is "roll high enough to win." Ideally the gameplay should exist above that, with the fortune representing a risk for players to evaluate in their declared move, instead of the inevitable arbiter of success.