Add documentation and added improved inertia alignment

[charnley]

No description provided.

[hagenmuenkler]

Looks good to me, I like the new readme, nicely written. I left a comment on the get_inertia_tensor function, where there is still some conflict of old and new code.
For that method, I would also put in the docstring that in calculates the inertia tensor around the center of mass. This is what everybody does anyway, but in principle you can calculate the inertia tensor around a different point.
In the readme, I'm missing an explanation that a full calculation (brute-force) of the minimal RMSD is possible (figure out all ways to match the mols, calculate optimal rotations for each); and that this is implemented in rdkit. So the point is probably to provide something that is much faster (especially for larger molecules) and still a good approximation.