Now that you mention it, I believe there probably are symmetries connected to these conserved quantities. They don't follow from the usual Noether's theorem because the "rate equation" describing the dynamics of chemical reactions is not derived from a Lagrangian. I proved a [generalization of Noether's theorem to Markov processes](https://arxiv.org/abs/1203.2035), but it doesn't cover the rate equation, which is nonlinear. So, I may need to generalize it further! Thanks for suggesting it.
> Is there a "free-energy" <-> time scale transformation? Where looking at a slower time is like looking a system with higher needed energy?
When you study chemistry quantum-mechanically you get the usual reciprocal relation between times and energies, related to the fact that \\(\hbar\\) has units of time \\(\times\\) energy. Here we are studying them phenomenologically, just writing down a rate equation involving some rate constants that come "out of the blue". There's more to say, but no time to say it!