I think the trouble is the vague Uber characteristics. They're left open for too much interpretation.
Offensive Characteristic
A Pokémon is uber if, in common battle conditions, it is capable of sweeping through a significant portion of teams in the metagame with little effort.
Defensive Characteristic
A Pokémon is uber if, in common battle conditions, it is able to wall and stall out a significant portion of the metagame.
Support Characteristic
A Pokémon is uber if, in common battle conditions, it can consistently set up a situation in which it makes it substantially easier for other pokemon to sweep.
General problems - what would be considered "common"? Something which occurs in 50% of all battles? 40%? 30%? And so on. Without any clear benchmark for common, this doesn't really give us any picture. I could argue Pokemom X has to risk Spikes, Toxic Spikes, Stealth Rock, and Sandstorm in switching in, meaning Pokemon X isn't Uber, and quite rightly in the current metagame, you'd point out I was being absurd. But where is the line drawn?
Secondly, where are these common conditions? Imagine a metagame where OU consists entirely of Stealth Rocks resistors, and the number of potentially useable Rapid Spin users is much, much higher. Stealth Rocks would be used much, much less, to the point where it may not be considered "common" (assuming we've found our benchmark for common). However, let us say something like Platinum happened, and we got a Forme which was a super powerful Fire/Flying, Pokemon Y, that we wanted to test for Ubers. Before the introduction of Pokemon Y, Stealth Rocks was an extremely rare move, but due to Pokemon Y's large weakness, usage will almost certainly rise, possibly pushing it above the common benchmark. When considering whether Pokemon Y is Uber, are we using the common conditions of the metagame prior to its introduction? Or the metagame after its introduction? The problem with using that first is that of course a metagame will react to the introduction of any powerful new Pokemon - introducing anything would likely change the common conditions to an extent, simply because this Pokemon Y will get used, even if it is merely Top OU and not Uber. The problem with using the second is that the metagame may get so centralized around defeating Pokemon Y, that Pokemon Y can't immediately be seen to be broken - it might hover at around 30% usage, but because the entire rest of OU is centralized into being able to cope with it, it can't rise further. Obviously, using these common conditions would be unfair, as they've specifically evolved in order to be able to cope with Pokemon Y? So if both sets of "common conditions" have flaws, and neither one is specified, then the term is vague to the point of being useless.
I hardly need to point the exact same problem applies to the term "significant" as it does to "common". How much is common? How much is significant? Is Pokemon X UIber if it can sweep 50% of the metagame easily? 40%? Etc, etc.
Metagame is much like conditions in above, insofar as that which metagame is never defined. Is it a) the OU metagame prior to the introduction of Pokemon Z, or b) the OU metagame after the introduction of Pokemon Z. a)'s flaws are that of course the metagame will centralize slightly around the introduction of any new threat - as an analogy, imagine Tyranitar being introduced to a previously Tyranitar-less metagame. You'd see massive shifts, simply because Tyranitar is very threatening as a Pokemon, Uber or not. b)'s are that the metagame could become so centralized that Pokemon Z can't sweep a "significant" percentage of the metagame, because every other team and Pokemon revolves around preventing it doing so. Neither a) nor b) is fair - a) would of course undergo changes, and in b) the effects may be masked.
Then, of course, it is another vague quantifier, "little". How much is little? From the moment it switches in? Given one turn to set up? Given two turns to set up? Given three? And so on. How much team support is defined as little effort? Would it still be considered little effort if it could sweep significantly with total ease, but required the entire rest of my team to become highly specialised in order to support it, and among ordinary team layout, was fairly useless?
I don't think we can really get anywhere until this is much, much clearer, because at the moment people are just twisting it so that it can justify the banning / unbanning of almost anything.
And on a very slight side note, I notice the Suspect Characteristic seems flawed anyway? If a Pokemon could set up conditions which made another Pokemon capable of walling every single other thing in existance, then apparently it wouldn't be Uber, as the the Support Characteristic can only be applied to those which set up conditions to make other Pokemon capable of sweeping, which seems wrong. Shouldn't the Support Characteristic read:
Support Characteristic
A Pokémon is uber if, in common battle conditions, it can consistently set up a situation in which it makes it substantially easier for other pokemon to fulfil either of the other two Uber characteristics.
Talking of which, the Support Characteristic also has another flaw specific to it. How many Pokemon is other Pokemon? For example, Pokemon A under by itself isn't Uber in the slightest. Pokemon B has the potential to provide support for Pokemon A to the point where Pokemon A is considered Uber, however, Pokemon B also provides valuable support for other Pokemon, without making them Uber. If we removed Pokemon A from the metagame, we'd have removed everything that could be considered Uber, however, Pokemon A by itself is not Uber. If we removed Pokemon B, we'd have removed everything that could be considered Uber, however, as Pokemon B actively contributed towards making other Pokemon more useful, you could have damaged the metagame in the process by removing Pokemon B who would have been very useful with the existance of Pokemon B. Here, it would seem better to remove Pokemon A, but this is an extreme example. What if Pokemon B made two Pokemon Uber? Three? Four? But in both these cases, it added three/four useful other Pokemon to the metagame? At what point does it stop being the Pokemon who become Uber, and start being the Pokemon that make them Uber?
Then there's the old problem with "how consistent is consistent" and "how substantial is substantial" but I dob't think I really need to go through that again.
This was mostly just a mindspew, but hey, it's 10:00 in the morning and I have free time to burn. Feel free to shoot down.