Revisit documentation recommendations for alignment

[handwerkerd]

Summary

The documentation currently recommends running tedana before doing distortion correction or non-linear spatial alignments to a template and then has a note saying the other option is ok, but deserves more testing. Both options are ok and I'm not sure if either is better so we should present both in a more neutral manner that described the benefits and problems with each

Additional Detail

Relevant documentation section:

https://github.com/ME-ICA/tedana/blob/b11d2547837d719b84b34c650b8d15c7f465ace8/docs/multi-echo.rst?plain=1#L461-L479

I think all the relevant points are there, but I'd reorganize to explain the benefits and problems with each approach rather than recommending one and making a note about the other option. (Full disclosure: It looks the way it currently does because of my previous recommendations and others probably wouldn't have recommended both in earlier versions of the documentation)

Tangential issue

An entire section of multi-echo.rst is not rendering on readthedocs:

https://github.com/ME-ICA/tedana/blob/b11d2547837d719b84b34c650b8d15c7f465ace8/docs/multi-echo.rst?plain=1#L443-L458

I think there's a missing blank line that's causing the problem. This is fixed in the rendered docs for #970 (yay strict linting!), but figured I'd note this.

Next Steps

• Have a few of us agree this is a reasonable change
• Edit the text

[tsalo]

It's also worth thinking about how dynamic distortion correction (e.g., with MEDIC) fits in. With dynamic distortion correction, it makes more sense to run the distortion correction before motion correction, so there's no real way to run tedana on motion-corrected-but-not-distortion-corrected data.