It's not that you reach a certain fraction and the reaction just suddenly stops. It's that the reverse reaction also occurs and, at equilibrium, the forward and reverse reactions occur at the same rate.
Actually, I wasn't even referring to an equilibrium reaction. Maybe I didn't communicate the issue well enough. I'll try again:
Consider: A + B
C
at a particular temperature of A, the molecules don't all have the exact same kinetic energy - some have it higher and some have lower. The ones that are sufficiently kinetic meet the energy demands for breaking bonds (reacting); their KE (kinetic energy) is equal to or greater than the activation energy. Once the "reactable" bunch reacts, the reaction is complete. Unless:
1. heat is provided throughout the course of the reaction to get more of the A molecules into the reactable state
or
2. Molecule A, before letting it react with the B, was driven to a temperature that ensured all or most of the molecules were over the energy hump. Only then it was added to B to start the reaction (this way all of the A molecules had the necessary energy prior to introducing them to react with B)
The reaction constant does not change as the reaction progresses if you are at a set temperature (and I imagine pressure, but that affect would likely be insignificant for nongas phase reactions). The rate constant is a function of temperature and only parametrically depends on collision frequency, activation energy, etc.
You got this right. The key phrase is "set temperature".
My point is this: if you set the temperature at an "edge" temperature (meaning a temperature at which only a fraction of the molecules posses the required reaction energy), then only half of those reactants will react. As the reactable reactants turn into products, the average reactant temperature goes down (because the only reactants remaining are the ones that don't have enough energy to react (as opposed to the start of the reaction when high energy molecules were factored into the average)).