That would require foreshadowing of the values of the variables, which means constantly looking back and forth at the known data which in my opinion can be distracting (at least for me) when your mind is working on a different qualitative realm (imagining atoms bouncing around, atomic bonds...