What would you like to do:
$> grep ixi.*328 .datalad/crawl/phenotype_mri.csv
ixi,1,328,61.78507871,control,,a-levels,white,,,married,retired,,Female,ixi,homo-sapiens,,https://files.osf.io/v1/resources/5h7sv/providers/osfstorage/5839b9fd594d9000461f6ce5,1,yes
ixi,1,328,62.08350445,control,,,,,,,,,Male,ixi,homo-sapiens,,https://files.osf.io/v1/resources/5h7sv/providers/osfstorage/5839b9fd594d9000461f6ce5,1,yes
$> grep ixi.*328 _stats/content-duplicates
b2394fb80018d293a0eab83023ab4d72 ./ixi/sub-328/ses-1/IXI328-HH-2295-T1_rep-0.nii.gz
b2394fb80018d293a0eab83023ab4d72 ./ixi/sub-328/ses-1/IXI328-HH-2295-T1_rep-1.nii.gz
it is the same content since the same url ;-) but metadata differs drastically -- I guess subject changed gender jumping 62 birthday ... I guess could happen, but not very likely. I haven't done yet full analysis of some kind to identify the conflicts but I guess it might be beneficial to provide some data cleanup, e.g. to remove data with conflicting metadata?