I had a game once that was quite the amazing comeback. I was playing Kha'zix mid lane and got pretty fed, but everyone else on my team was feeding. At 22 minutes, they were 12,000 gold ahead and on our Nexus turrets, and surrender votes were being thrown. Me and some other guy saw that they were starting to get really cocky, thinking that the win was assured no matter what, and were letting themselves get caught and throwing. We kept hitting no to the surrender votes. We kept the game alive, got different objectives as we could instead of teamfighting, and caught back up. At 50 minutes, we won with a 6,000 gold lead. It was a very fun and satisfying game.
As for your system, while I do think it would be good, I also think it would be impossibly skewed and near impossible to code. Think about it. Matchmaking based on a % of the times players hit no to surrender votes. Players with fewer surrender votes to hit no to will always be matched with quitters, even if they aren't quitters themselves. Players with more surrender votes that they hit no to will always be matched with other players with a ton, so they'll never find another surrender vote being thrown to keep them up in the system. And, because there are more quitters than optimists at low ELO, low ELO optimists will be given very long queue times. Not to mention, all the quitters piled on one team against a bunch of other quitters. Not only do quitters give up easily, but they're also easier to tilt. So, whoever gets first blood basically gets a free win as the enemy team spirals into an avalanche of tilt and then surrenders at 20. While I think the system would be good in helping optimists be matched with other optimists and basically make their games easier, you also have to think about it from the quitters' perspective, which has their games being far, FAR more toxic and unfun.
Then there's the fact that this is also basically the current ELO system. The optimists often end up rising in ELO faster than the quitters, and as such get to higher ranks easier. If they have the knowledge to carry a game, even with someone who gave up, it's not hard for them to. Low ELO is mostly quitters with a few optimists because the optimists climb, while the quitters don't. High ELO is mostly optimists with a few quitters because their optimistic attitude just helps them climb. Plus, having an optimist in your game can raise moral a ton over if it was just a bunch of quitters. I've had games where I carried, I've had games where I got carried, and I've had games where we surrendered and I sat there questioning why because I was far enough ahead to carry.
