[Freesurfer] FSL and freesurfer: combine runs in volume or on surface?

Dear all, I am trying to figure out how to bring my fMRI data from FSL to the freesurfer surface.

The way I understood it from the tutorial "3.0 Surface-based Group Functional Analysis"( http://surfer.nmr.mgh.harvard.edu/fswiki/FsTutorial/FslGroupFeat)

There are two possibilities of combining multiple runs per subject: - using mris_preproc for each run and having runs as frames in the lh.copex.mhg file - running first gfeat in FSL and registering gfeat copes to fsaverage ( Runs combined in a GFEAT within a Subject), which would result in 1-frame lh.copex.mhg file

I wonder if it is possible to say which procedure is better/more sensitive? Should the fmri data be non-smoothed before this step for both above options?

Thank you! Cheers, Aga