Implement support for moving sources and dust (and associated Lorentz boost to emission)

[astrofrog]

We should more clearly lay down what assumptions we are making here. In particular, we should technically be updating the opacity the photon sees in each cell every time it enters and escapes from the cell, but in practice this will likely never happen. Except of course if we start dealing with gas. As long as we are only looking at dust, then the opacity doesn't change enough in wavelength to be worthwhile. So we should make sure we implement only what's needed for moving sources and featureless moving dust, and we can later consider whether to use update_optconsts for every cell (when the opacity has features)

[astrofrog]

Things that remain to be done:

• [ ] investigate raytracing
• [ ] write up simple tests
• [ ] write up example models

[astrofrog]

At this point, the bit-level tests pass locally

[astrofrog]

I'm going to close this, as I will no longer realistically have time to pursue this in the future. However, if anyone stumbles upon this and is interested in helping push it forward, please leave a comment!