Truth: how much do sheets and other game materials support this fact?
Mechanical Support: how much do game mechanics support this fact?
Valence: how much does this fact meaningfully help the character achieve her goals?
Impact: how much does this fact meaningfully affect things people care about in game?)
Truth is how much a fact reflects reality; I'd say, specifically, the game's explicit background. Weak Truth would be something like your character sheet saying you're the world's greatest swordswoman, but this not being substantiated anywhere in game. If you make enough effort and have enough Mechanical Support, maybe you'll convince game you're the world's greatest swordswoman (and the consensus reality of the players /is/ the reality that matters most), but it still feels a bit tacked on. Normal Truth would be having it as your player list role and on the sheets of people who know you, or maybe the swordsmanship greensheet. Strong Truth would be something like your skill with the sword having saved the town pregame and being a key part of the backstory, or at least have it be a nontrivial part of the scenario. This means everyone knows this fact, and suggests that people care about it as well. (The weakest possible Truth would be the GMs being all "of course she's the world's greatest swordswoman!" but neglecting to write it anywhere in game at all.)
Mechanical Support seems pretty obvious. If you're the world's greatest swordswoman but you have an average CR and no cool combat abilities, that's weak Mechanical Support. (This is more common with things like "world's greatest cook" in a game with no cooking mechanic, say.) You might intend having the highest CR in game to be reasonable Mechanical Support, but without thought your mechanics may not end up giving the effect you want. It's great in a martial-combat heavy game or one with an important dueling mechanic, but if the only reasonable way to fight in your game is with guns, having the highest CR is weak mechanical support. (Though, if the theme is that technology has made your skills obsolete, that can still be cool.) WTF's strong Mechanical Support is the ability to create seemingly-unrelated mechanical effects with your fact, which is probably more than you want in most guild games. But if your game has alpha-dominance struggles where you can win people's loyalty by beating them in a swordfight, that's strong Mechanical Support. The main income being random drops you get from killing monsters is pretty good Mechanical Support as well. Mechanics that depend solely on player skill generally lead to weak Mechanical Support for any character supposed to be notably good or bad at them, though careful casting can give a kind of pseudo-Mechanical Support here.
Valence in WTF can be stated as "the ability for something to help you achieve your goals". If you're the world's greatest swordswoman, but this isn't useful because people don't die permanently and all your goals can only be solved by committee plots, say, that's weak Valence. If your goal is to kill all the vespid and your attacks can do this easily, that's strong Valence. Most people have some goals their skills are effective at and some where they're less so, especially in a PvP plot where the GMs try to balance the sides. Things can have Valence without Mechanical Support: if you're the President of the United States, even if you have no cool abilities, there's a good chance you can roleplay authority well and achieve goals that way, given player competence, especially if your Presidency has strong Truth. Even without strong Truth, though, some players are good at using roleplaying to create their own Valence. But, generally, you do want there to be a path by which a character has a reasonable shot at achieving their goals.
I think Valence can be generalized to something like Impact which tells how much something meaningfully affects the game. Regardless whether it's a good or bad outcome for anyone in particular, someone should care about anything you put into game. For example, one sort of thing that's common in Guild games are "horrible secrets" with weak Truth and no Impact. If my goal is to make sure no one finds out I'm secretly an android, but this isn't substantiated outside my character sheet and there's no reason that my secret getting out hurts me, that's pretty lame.
These can apply to things like goals, not just abilities. If I have a plot to make a widget, but my desire to make the widget has weak Truth (I don't have any convincing reason why it's important in my character sheet), the widget has weak Mechanical Support (doesn't do anything useful) and no one else cares (minimal Impact) it's likely to get dropped on the floor. If the widget is key to someone else's ritual, or my company gets more resources based on how much science we do, that can give it more Impact. But it's probably best to make sure that goals have at least moderate Truth if you don't want them to look tacked on and get easily dropped.
(Side note: my spellcheck is sexist and kept autocorrecting "swordswoman" to "swordsman" in this. Grrr.)