Parallelisation of class source term calculation

[robertsawko]

I believe that at the moment all of our MPI processes are repeating the job of calculating all source terms. This has to stop!

[robertsawko]

Our code works rather well for parallel runs. I think it is actually parallelised thanks to forAll cells, but I can't say for sure. This is no longer an enhancement, bur rather a question.