Regarding the video transcriptions, I would imagine you guys tried this, but YouTube was pretty good about getting close with its captions, then they can be edited later. I helped my wife with some demonstration videos she had made and it took us maybe 2-3 times the length of the video to fix up the captions. Sometimes it was spot-on and sometimes it was hilariously wrong, but it was still better than doing it all from scratch.
If that’s still too much then you might be able to export the text from YouTube’s first pass, then you could just drop that in the git repo that was discussed to give a good starting point for any helpful souls who want to improve the transcriptions.