I think it sounds like an excellent “deep dive” project, but don’t get your hopes up that a million people will use it. Ultimately I believe this will be a test of your statistics and math skills more than anything,
I think you could probably reasonably complete the application itself in 1-3 months, but I have no idea how long it would take to complete the matching/recommendation algorithm.
Hey, not to discourage you at all, but from what I can tell most viewers are a bit bored by watching episodic Factorio videos. I would recommend creating longer-form videos that encompass your entire challenge rather than an entire series of it. It’s a lot less video output for the effort you put in, but you have the advantage of being able to narrate the entire challenge as a story. If you want some good examples of this, check out Dosh Doshington’s content. He doesn’t post too often, but when he does it always does well.
It might be worth giving Helldivers 2 a try, it may not end up being his thing but it has a good few milsim elements