I think we all need to go back to basics here.
When it comes down to it, there are two main styles of play when designing a pokemon team - Offensive and Stall. (I know this is a broad generalisation but it's effectively true - trying to strike a balance between offense and stall is of course possible but doesn't actually affect the relevancy of my next point and would just make it about 3 pages longer to explain in the long run.)
When you choose a style, you have pokemon which are the best in that style. This is indisputable, it's in the basics of the game and the competitive mindset that you don't want to use something that's worse at doing the same job than something else you have available.
Mence is one of the best attacking pokemon in OU. Again this is indisputable and people from both sides of this debate will agree with this. Latias was also one of the best offensive pokemon in OU. And now the discussion has moved on to Scizor, who is again, one of the best attacking pokemon in OU.
If all 3 end up uber then people who favour and offensive style of playing will be turned off the game.
We seem all too happy to ban the offensive threats but things like Blissey who can survive any special attack in the game (with a few exceptions, and even then it might use protect) hasn't even been considered for a ban.
I'm going to link to sirlin's 'Playing to Win' articles again (which IMO, should be read by everyone who posts in this topic)
http://www.sirlin.net/ptw-book/what-should-be-banned.html
Criteria of a Ban
A ban must be enforceable, discrete, and warranted.
Now, enforcable and discrete are no worries here, that's pretty obvious, the category I'm going to look at is 'warranted'
(I'm not quoting the entire passage because it's massive and a lot of it is about bugs in games, so I'll just get the important bits)
Warranted
Here is the whole issue, of course. If it isn’t warranted to ban something, we don’t need to even consider whether it’s enforceable or discrete. The great lesson of competitive games is that hardly anything warrants a ban.
I don't fully agree with this in the context of pokemon, some things DO need to be banned. But not EVERYTHING
How does one know if a bug destroys the game or even if a legitimate tactic destroys it? The rule of thumb is to assume it doesn’t and keep playing, because 99% of the time, as good as the tactic may be, there will either be a way to counter it or other even better tactics. Prematurely banning something is the scrub’s way. It prevents the scrub from ever discovering the counter to the Valle CC or the diamond trick. It also creates artificial rules that alter the game, when it’s entirely possible that the game was just fine the way it was.
Does mence have counters? Answer - Yes.
It also usually leads to an avalanche of bans in order to be consistent with the first.
And that's the - 'lol, scizor is only played b coz mence' argument countered in one, if you ban one thing it WILL have a knock on effect. Remember that mence is only being considered because of the Garchomp ban.
When players think they have found a game-breaking tactic, I advise them to go win some tournaments with it. If they can prove that the game really is reduced to just that tactic, then perhaps a ban is warranted.
So as it stands, nobody can prove that the entirety of the OU metagame is reduced to 'Salamence or lose'
If the usage of mence ever gets close to 100% the ban should be considered. As it stands it isn't anywhere near that, its not even the most used.
I'm going to end with another quote which I think nicely sums up my entire argument.
The only reasonable case to ban something because it is “too good” is when that tactic completely dominates the entire game, to the exclusion of other tactics.