The critique of "Hundred" is very simple. The geography and number of players combine to present a terribly short range of options for potential outcomes.
The map has seventeen supply centers and a nine-center win condition. As anyone beyond second-grade arithmetic can tell you, seventeen is an odd-number; therefore, splitting the centers two ways is impossible. Thus, the game eliminates the possibility of a two-way draw; the closest split is 9-8, but obviously this means one player wins.
So the only options now are a three-way draw, a win with one survival and one elimination, and a win with two survivals. But since the game starts with three players, a three-way draw is ill-advised. What is the purpose of playing if you will end up drawing with all parties included? For the same reason a seven-way draw is undesirable in the classic map, a three-way draw is absolutely reprehensible in Hundred. The game might as well be cancelled, though I suspect we'll see players draw for the sake of completing games. (As they should. I suspect the statistics will prove this critique stronger.)
Removing the three-way draw, then, we see that a win is the only possibility, with the only choice being an elimination or two survivals. The only logical solution for the players who are out of contention for winning is to play for two survivals. The rationale for this argument is the same one John Rawls uses in his arguments concerning the veil of ignorance. Because the players ultimately do not know if they will be the one eliminated or the one who survives, they should play to have all other parties survive. That way, if in the one-survive-one-elimination model they were destined to be eliminated, they have improved their stock by surviving; and, if they were to survive, their position is no worse affected by the survival of the other party.
Given all of this, the Hundred variant is a spectacular waste of time because there should only be one result at all times: a win with two survives. This is because, as soon as one power's chances of winning are inevitably ruined, that power will always actively seek to carry another power to a victory as quickly as possible to ensure his own survival; and, in the face of this alliance, there is no diplomacy, as the winner cannot be dissuaded from his inevitable victory nor can the junior partner be dissuaded from surviving. The third player can then continue trying to win, but unless his tactical position is extraordinarily sound or he is significantly more skilled than the other two players, the leading power's victory is almost inevitable.
For a case study in this, see the public press live game that Candelarius, raapers2, and I just finished: http://vdiplomacy.com/board.php?
gameID=418 In it, I was England, and consecutive stabs by France and then Burgundy made it impossible to win. I therefore had to resort to throwing the game as quickly as possible to avoid being eliminated. I tried to throw it to Burgundy, but Burgundy spurned my advances; so, I went to France, who happily agreed. The result: France, 9-center win; Burgundy, 7-center survive; England, 1-center survive (in bloody Scotland, no less!). After Burgundy stabbed me the very turn I tried to give him the win, I literally had no diplomatic option BUT to give France the win. I could not play for a draw, as discussed before, and the only player willing to take the win on a silver platter was France.
Hundred discourages the win-draw-survive paradigm that has come to be THE set of victory objectives in Diplomacy, instead encouraging the very vassaldom and unbreakable alliance syndrome that is counter to the spirit of Diplomacy. I therefore propose that any true lover of the hobby avoid this variant until it is further modified. I know I will be.