Long post incoming on my analysis of the uses of an endgame and what makes them exciting:
I don't think its quite fair to say that this year's end of match is more thrilling than 2013's endgame, and then extend that statement to endgames in general. The consensus seems to be that it was far more gripping to see last minute truss shots, high goal shots under CRAZY defense, or misses in the last 5 seconds than the 2013 endgame, which consisted of 2 or 3 robots that would simply latch onto the pyramid. The 2013 endgame was un-entertaining because it interrupted the play of the match simply so that both alliances would almost automatically get an extra 20-30 points, which very rarely swung any matches. Of course some teams were quite inspirational with the pyramid (254, 148!) but in reality those teams seemed almost non-existent. Even the final match of Einstein ended in 5 robots hanging, which really only padded the scores, and had no element of suspense.
That being said, I don't think this is necessarily true for all endgames. I particularly enjoyed the 2012 bridge balancing endgame, as it added a significant amount of suspense via an action that was significantly more challenging than hanging on the pyramid, required fantastic interaction between alliance members, and was worth enough points relative to the rest of the game to be significant in outcomes.
Finally, I would also say that the need for an endgame depends on the rest of the game: I think an endgame in 2012 was a necessity to break from very routine collection and shooting of balls. For example, without the endgame the blue alliance in the 2012 Einstein final match (
https://www.youtube.com/watch?v=K84uckmXg_c#t=1122) would have had very little hope once they missed shots in auto and were behind. However, the 2012 endgame gave the blue alliance a reasonable chance to win the match by balancing: it seems plausible that the red alliance could have messed up the balance.
In contrast, the 2014 game in my opinion has absolutely no need for an endgame, because a 3 assist cycle ball is worth so many points that any given alliance can jump into the lead by scoring a 3 assist cycle and preventing the opponents from. Last second completed cycles and trusses were game changers this year.
I suppose I come to 2 conclusions:
1) The question of whether an endgame should return is dependent on the rest of the game, and I don't think it makes sense to talk about it out of context.
2) When an endgame is implemented the GDC needs to make sure to balance many factors of it: the difficulty of doing it (you don't want it to be free points like 2013, but it shouldn't be too hard to do, like a 30 point climb); cooperation necessary between alliance members to complete the endgame (if one team failed to hang on the pyramid, you only missed 10 points. If one team messed up on the bridge, you would lose the match); and the relative point value of the endgame (it needs to be enough to swing some matches, but shouldn't be overly domineering).