De-fluidize material data for porous media flow

[tulioricci]

Mostly as a reminder, we should think of a way to untangle material properties to fluid properties for porous media flow following review of PR #875. Currently, it is all stashed in gas_model and fluid_state since it is passed everywhere around in the code. Although this contradicts the naming, porous media flow requires a "porous media", we can refactor the current proposed structure for organization and clearness,.

Future physics-targeted experiments+simulations will have to go deeper in this topic. Thus, not-so-distant extra implementations will likely be a good point in tackling this.

@majosm @MTCam