I suspect that this line is an important bit:
It has been turned into something that can easily be scored.
A handy metric is one of those things so useful that you will start by being tempted to do a great deal of epistemological violence; and be lucky if you end up even being able to recognize when you are doing so. But did I mention useful(albeit with the risk of enabling you to do the wrong thing with speed and certainty)?
I’m also reminded of D&D, and other tabletop RPGs’, battle against the forces of min-maxing munchkins and soul-devouring rules laywers. Nobody expects a ‘realistic’ simulation to be viable with pencil, paper, and a handful of D20s; but the tone of the game changes sharply when people shift to gaming the rules rather than playing the game; and there are basically no known architectural solutions: some rulesets are more blatantly exploitable than others; but they all end up relying on a ‘GM doesn’t have to put up with that nonsense’ clause(and the existence of a GM who won’t put up with than nonsense; but also isn’t so capricious that players feel that blind adherence to the rules is the only alternative to rocks fall; everyone dies style management).
And the ‘existence of a GM’ bit is the real kicker. You can iterate your way to somewhat less broken rules(and that’s a good idea; broken isn’t good); but if you can’t or don’t trust some amount of benevolent discretion you have to rely on going by the book rigorously enough that gaming the rules becomes a viable, typically overpowered, strategy.