Reproducibility doesn't require the original data set to be present, only that the method of conducting that particular study reproducible. There are huge differences.
I probably sound like a broken record, but this is standard across scientific disciplines. No raw data and people call bullshit.