That's a great point. It's similar to why people often set up their problems as 3SAT or convex optimizations--you get to benefit from all future improvements to 3SAT solvers or optimizers.