Including all growth and electrical/chemical effects is not yet possible, because they are not yet all known. Furthermore, knowing all of them would probably tell us that there's no such thing as a generic neuron model; each cell has a very specialized gene expression profile that near-optimizes it for its role in the network. It may be that this can be reduced to expressing a neuron as a function of some hundred-odd parameters (as a guess), but it is not yet clear what the controlling parameters are.

The MURG question of computational modeling of the human brain is a very open problem with no solution in sight. I was not familiar with them, but they do at least seem to be composed of people with useful technological skills. It'll be interesting to see if they come up with anything cool.