Demonstrate the effects of the perturbed polybeam through imaging

[steven-murray]

Maybe @r-pascua wants to update this with details...

[r-pascua]

Right. So the current polybeam tutorial notebook shows a gridded version of the beam and demonstrates its effects on the data by plotting an autocorrelation amplitude as a function of beam rotation angle. I was thinking that it would be nice to have some comparison images: put something on the sky (like the HERA text from the pyuvsim reference sims), simulate visibilities with various perturbed polybeams, then image the simulations and show how they compare to images made with the unperturbed beam simulations. This is by no means a priority, but I think it could be helpful from a pedagogical perspective.