Chemistry solves are not guaranteed to solve for the entire time step

If you look at this code

https://github.com/ESCOMP/atmospheric_physics/blob/0f2e4afd1a54b41dc6d8529b9c25281114c87a7e/schemes/musica/micm/musica_ccpp_micm.F90#L243-L252

we call solve once. That returns something called the [solver stats](https://github.com/NCAR/musica/blob/6b2952c2f3bb56c02f3ee32a797f308177b90d81/fortran/micm/micm.F90#L18-L27). One of the attributes is the `final_time`, which represents the amount of time we actually solved for. So if you pass in 60 seconds, we might return 45 seconds if there is an especially stiff system. We need to repeatedly call solve until the full integration is complete.

Mostly, we will solve for the entire time. It is rare that we don't solve for the full time, but best to be correct