(As a note, I fully agree “Score” shouldn’t be based solely on Kills & Assists; this post is about K:D as a method of judging in-game effectiveness.)
As people have said, the K:D ratio is pretty hard to apply any value to across the board because the Kills score only really indicates effectiveness on certain characters. I think it’s absolutely valid to expect Marquis to have a solid Kills score, much like Reyna should have a solid Assists score and Oscar Mike should have a solid Minions Sent To Fiery Robot Hell score. Different characters are awesome at different things, and that’s part of what makes Battleborn rad.
But while the value of Kills varies greatly, Deaths on the other hand strikes me as a much more consistent measurement than Kills or K:D. Because some characters should be getting a big K, but nobody wants a big D.

No character has been designed with death as a necessary or encouraged part of their kit, and there’s no role in the game that is helped by dying: even tanking is a role about mitigating incoming damage to prevent deaths, not dying for the cause. Whatever your role or chosen playstyle is, you can’t do it from post-death spectator cam.
Deaths are gonna happen, certainly: finishing a game with a handful of Deaths on your card just means the enemy team wasn’t massively outclassed (or asleep
). And sometimes you’re on the receiving end of that massive outclassing and just get absolutely torn apart; it happens. But if it’s happening all the time and you’re racking up a dozen-or-so Deaths per game, then I guarantee you’re taking unnecessary risks, or not paying enough attention to your surroundings, or not learning from your mistakes, or not playing to your strengths, or your character’s, or something. And I’m not an elitist at all when it comes to online gaming, and I’ve never abused someone in a video game, but I will absolutely use those 16 Deaths on your sheet to silently infer that you weren’t helping much. 
I’d probably be able to confirm that with the rest of your stats during the post-game, since it’s hard to pull good numbers on the other statistics when you spend half the game waiting to respawn, but at a quick scoreboard glance I do think Deaths by itself provides a much more accurate assessment of effectiveness than Kills or K:D and actually maintains validity across the roster.