Interesting approach.
I created a database of women when I was dating more (a couple of years ago).
I had 15 different selection criteria. I tweaked the criteria as I went along. I was going to add weighting factors but then I stopped, realizing that I was obsessing more over the design of the dbase than the women LOL.
But I didn't think to go back and enter historical info. In a less scientific approach, my mind figured out the types that would work for me and those who would not.
I concluded that I was becoming Pygmalion, designing my ideal woman..but who remained just an ideal.
It wasn't that I was looking for a supermodel / rocket scientist / nymphomaniac (some of them might have been a bit out of my league ;-). I had things, traits that I liked that were kind of mundane.
One takeaway for me was that I could reliably figure out what someone was like with very few cues and with very little time spent, that is, stereotypes were fairly reliable.
But if you see some qualities in everyone, you can probably get to the point where you can be with many more than you thought possible.
If you can let go of the "happily ever after" expectation and just be content to enjoy company, you might find that your heart's content is no further than your own backyard, so to speak.
Maybe we need to let serendipity have her way. Maybe we have too much freedom of choice. Maybe something short of arranged marriages would work.
But you have at least inspired me to play with my dataset and dbase and see what it can produce.
How romantic!!!! LOL