You know that if that position had been reached in game 3, Carlsen would have played on. But in a game 12, it would be foolish for the champion to take risks when a draw would lead to a tie-break in which he is the favorite. Engines, commentators and even the players agreed that black was better, but it's still pretty unclear for everyone involved how black could have turned that advantage into a win. If the path is unclear, it is, by definition, risky.
This 12 draws result raises a lot of questions on the state of chess and the format of the contest.
GM Judit Polgar felt like we reached a climax of the event at game 10, but with only 2 games left, stakes were too high and it resulted in two kinda lame draws, something that might have been avoided in a longer match (18 games is the number often mentioned).
In a double edged observation, GM Alexander Grischuk noted that it was probably the most accurately played world championship match ever (it's probably something like 0 blunders and 8 inaccuracies total over 12 games!) and wondered if it mean it should be played in a slightly more stringent time format (so players make more mistakes) or even use things like Chess960 (a chess variant invented by former world champion Bobby Fischer where the configuration of the pieces on the back rank is randomized, making it nigh impossible to work on openings and known lines and so enforcing creativity and chess IQ). Obviously, the rebuttal for the first idea is that you want both players in a World Championship to play their best possible chess and the rebuttal for the second is that it is a bit absurd to use a chess variant to crown the best chess player, unless the variant becomes the main way to play chess.
GM Peter Svidler noted that while proclaiming chess itself is done or broken would be an exaggeration, a World Championship that ends in 12 draws is not a great publicity for the sport, even if many of the 12 games were interesting for chess enthusiasts and experts. He brought forth an idea that has been floating around, which is to play the tie-breaker before the actual competition, to create an imbalance from the get go. The rebuttal is that the imbalance might be too great and affect the games too much, but I guess the rebuttal to that rebuttal is that several World Championships have been played where the challenger had to beat the champion to become the champion, drawing over the serie meant the champion kept the crown.
GM Anish Giri felt the whole debate is over reaction since one or two moves played differently in game 1 could have resulted in a Carlsen win, changing the whole dynamic of the match. So, basically, running for the hills because a couple guys did not manage to capitalize on their advantages over 12 games is a bit much.
While his pedigree is a lot less impressive that the people I mentioned above, youTube chess commentator Agadmator, made a point that I found interesting. World Championship matches always have been contests with a tremendous amount of preparation, but in the past it was humans studying human lines analyzed by humans, so even main lines could contain inaccuracies or had flaws that could be found and exploited over the board. Now, humans are studying computer lines analyzed by computers that contain zero inaccuracy. In the eyes of Agadmator, that leads to games that are played with extreme precision while both players are still in their preparation, but once they are out of it, it leaves them with pretty sterile positions. Unable to continue to play perfectly once left to their own device, the quality drops and players are just content to play solid safe moves, resulting in draws.