Whether or not K/D matters is determined by the particular game/gamemode in question. And even then, a person's K/D might be considered irrelevant when their total gametime is added to the equation. Someone who has a low K/D might actually be a good player, but their K/D is low as a result of them not understanding the gameplay mechanics during their first few matches (assuming we're talking about a purely competitive game). Back when I first started playing Battlefield: Bad Company 2 (my first battlefield game), by the time I fully understood the different aspects of the game, I had a .77 K/D ratio. It took me a good while to work my up to my final K/D ratio (Battlefield 3 was released) of 1.22.
Now taking gamemode into consideration; K/D can mean a lot, or very little. In team deathmatch, K/D is everything. In free-for-all, not so much (you can go negative and still win, provided that your deaths are spread out amongst the other players). With gamemodes centered around objectives, K/D can be considered a factor, but only when compared to the player's ability to complete the objectives. Using Control from Destiny as an example: Kills give points, but flag control can mean everything (due to the score multipliers). If a player (we'll name him Bob) has a negative K/D for that match, but he was always able to ensure that his team was in control of the majority of flags, he could be considered a good player as he's ensuring that his team has the capability to make up for his poor K/D (assuming that they are competent players). In this scenario, Bob's K/D might be poor, but he's most certainly pulling his weight.