Salamence takes out the enemy at the cost of its own life, AND it has no control over what it kills.
Well, this theory is flawed when your opponent will switch to thier "check" when Salamence is sent out, the person on the recieving end has a choice of six (or less) pokemon to switch in. That pokemon will likely die (given other calcs run throughout this thread) this is most likely the same thing people use to stop a DD mence (as they have little knowledge to the set should this be mid-game) which those same pokes are often used to stop T-tar ect. While the Salamence user has no actual control over the opponents moves likely something imprtant will die, even so no wall-breaker can have complete control over what they kill, as the opponent can switch in any of thier 6 pokes, But the real problem is Salamence has killed one of yours while you are forced to make an obvious move or be punished by another Salamence attack.
But possibly the worst part is this is a Metagame bent around stopping a Salamence, yet Salamence can still guarentee at least one kill and then puts you in a good position for the rest of the match.
Salamence takes 35% or 41% after switching in and attacking; that is easily enough to attack again and likely take another foe with it and has revealed at least 3 members of your team and has broken your defensive core and likely put your revenge killer in a poor position,
Even if you happen to have fodder, which is difficult to know during the early or Mid- game as you have incomplete information, Salamence has 100 base speed, a stall team is hard pressed to outspeed it or has to take another hit. (Btw salamence will not do 40% unless resisted and the Mixmence set has perfect coverage so he can easily use another attack type)
If Mixmence were a Yu-Gi-Oh card, it would read: "Sacrifice this monster to destroy one enemy monster, your opponent gets to choose which one."
The problem with this analogy is in Yu-Gi-Oh you have a deck of (not sure about this) 60 cards losing one is not as important as losing one in a team of six, while I agree that Yu-Gi-Oh is also a strategy game and some other analogies can be compared this one can't hold any water.
the game is and should be based on intelligent sacrifices and a war or attrition.
When facing a Salamence you have an~ 50/50 chance to either be swept (by a DD mence) or have a member of your team killed by a MixMence. Their is no room for any intelligent sacrifices when if you try to sac something you can get set-up on. This has happened to me before, I tried to sac my (fairly) useless lead and then was set-up on, and then I had to face Salamence that could kill most of my team and had to make a poor choice that likely lost me the game. I have done the opposite thing as well, switched in my Skarmory to wall what I could only guess was a DDMence and then was 2koed by Fireblast and had then lost my counter to other key players. Forcing your opponent into a game of guessing and then still having the upperhand afterwords is not intelligent gaming.
People rant about Salamence's power, but it's really easy to ignore the consistency and dependability lost by choosing to use a pokemon like Salamence. The reason why bulky offense (and stall at one point) was popularized in the first place was due to strong consistency-- which is a huge benefit to playing that style.
Salamence is consistent in getting AT LEAST one kill per game, and unlike other Wall-breakers like Infernape he can't be resonably walled, as very, very few pokes can take two attacks from a Mixed Salamence and then do ~67% in return, not to mention then be able to take an attack from a DDmence who could have set-up in that switching turn.
While I agree with most of your (very good) summing up of Salamence's two popular sets, Salamence can use both sets forcing a game of guessing on your opponent, and then has the raw muscle to power through even with correct (lucky) guesses, and can seriously punish wrong guesses.
@ Vile
Salamence doesn't have any thing that can switch into both sets, no other poke can say this without hurting their chance to do actual damage otherwise E.g. Focus Punch Gengar and U-turn, Earthquake, Thunderpunch Infernape.