Built a work flow and pipeline to take newpipe video data, convert to freetube data. Scrap project playlists I made for the transcripts of each video still up in that playlist, slapped a rotating tor nym/id system to avoid soft IP address bans by YouTube at between 3-10 video dice role, then take the transcripts and feed my individual projects information via this system. 1500 videos cataloged and all done locally.
With a lightweight history txt file with video titles in what playlist so when I do a bulk dump end of month I only pull new videos added from that specific playlist.
Never have to touch Google directly again.