I have uploaded a mulitlingual csv with 500 or so rows. Many rows have default translations. Now some work has been done by translators/proofers. I have about 400 approved. Now when I build and download translations, I get identical files - with all my original rows - whether or not I specify in project settings to “Export only approved translations.” Since there is nothing in the file to indicate the string’s Crowdin status, I have nothing that lets me know, once downloaded, what data is good. In my situation, without the ability to export only approved translations, I have no way to use the work that has been done in Crowdin.