First off K12 films can go through K14 with a prehardening step. And, K14 films can go through K12 processing.
The problem is that these two crosses will give bad results, as the developers and couplers are different. So, you would have to redesign 4 developers to get the right result. This is doable. It has already been done in experiments. So, there you are. All it takes is time and money.
The easisest approach is to expose 3 sheets of film to an RGB image in a beam splitting camera, develop them each in a separate Kodachrome like reversal process and then tape the individual separations together. This has also been done and it gives good results.