I dispute that. If you ask, even for a particular game, "what is the best 6-pokemon team for competitive 1V1 play" you will get a different answer from everyone. "Highest damage per turn under purely favourable conditions" does NOT somehow make that team "the objective best".
Just because (statX) or(stat1+stat2+stat3/limiting factor 1+stat4...) is higher than any other possible combination DOES NOT mean that it is somehow "objectively best" in general, though in a badly designed system it might be. If it is possible to make an "objectively best" setup in a given game that will always succeed in all situations, that is not somehow a universal constant, but is a failure of the designers to make a proper system. This is one of the reasons why I advocate more simulationist approaches than adding up stats, as some player will always say "build X is broken! All its stats added up are more than mine! NERF NOW!" Sometimes (in badly balanced systems) this might be true, but usually it is due to confusing "min-maxed for X" with "objectively best".
Simulation does NOT mean replicating our universe in detail. It means generating a simplified universe with internally consistent laws.