I ran the fmriprep with the default pipeline but the brain mask was cutting off. Then I re-extracted the brain by running fsl_anat, and then re-ran the fmriprep by adding the flag to skip skull stripping. The new output looks great, but in the html file the functional and structural alignment was displayed like this (the old FS output before I corrected it):
I checked the actual preprocessed output, and they looked fine. All the other preprocessed output (including the freesurfer structural output) looked good, so I think I should be all good here. But I still wanted to post this out in case the re-preprocessing caused any issues that I missed.