ENH: Add Neuromag phantom subject

I created an incomplete FreeSurfer subject directory for the Neuromag phantom that contains the head surface and CT-as-MRI T1.mgz. With this, you can do a plot for this example that would look like:

import mne
import numpy as np

data_path = mne.datasets.brainstorm.bst_phantom_elekta.data_path()
info = mne.io.read_info(data_path + '/kojak_all_200nAm_pp_no_chpi_no_ms_raw.fif')
sphere = mne.make_sphere_model((0., 0., 0.), 0.08)
trans = mne.transforms.Transform('head', 'mri', np.eye(4))
fig = mne.viz.plot_alignment(
    info, trans, 'phantom', bem=sphere, surfaces=('inner_skull', 'head-dense'),
    coord_frame='meg', show_axes=True, mri_fiducials=True, subjects_dir='.',
    dig=True)
fig.plotter.screenshot('alignment.png')

FYI for folks who haven’t seen it, this is what the phantom actually looks like:

Options:

1. Add to bst_phantom_elekta. I don’t like this because the phantom is more generally usable than this, and it’s not part of the data Brainstorm actually distributes, either.
2. Add to mne-misc-data. It’s small so this is doable, but not the best organization-wise.
3. Add a mne.datasets.fetch_phantom. It’s more stuff in that namespace (not great), but probably the cleanest because you can tell it explicitly which subjects_dir to put the subject into. It’s also easier to update/add files in the future and for users to get them, similar to how fetch_fsaverage works.
4. Don’t distribute this, nobody actually needs it (though I do, other MEG folks might, and it would make the bst example nicer!)

I prefer option (3), any thoughts?

[larsoner]

Or maybe fetch_phantom(kind='neuromag')

[larsoner]

(this could mirror the design of get_phantom_dipoles, which would be nice)