Even in things like applied math, it's pretty arbitrary which assumptions and formulations become standardized when tackling similar problems (and it usually comes down to the effects of prestige on your network/ 'reach': the more prominent people you get your ideas in front of, the more likely your work is to become the 'state of the art' default approach, and the more prestigious your university (and the professors you collaborate with), the easier it is to get your ideas in front of important people.