I really liked Enki's post. That's the angle from which we should be looking at tests like this.
All through Black and White as well as X and Y, we keep having suspects where people complain about how we are "forced to run obscure counters" and that the Pokemon in question "stifle diversity." Finally, we have a metagame that is pretty much as diverse as can be, with Pokemon ranging from OU to RU being legitimate metagame threats. This, in tandem with the Mega Pokemon that teeter on the line between broken and non-broken makes the metagame match-up reliant. And unfortunately, this is a side effect of a diverse metagame. If you want more Pokemon to be viable, you have to accept that there will then be more threats to consider while teambuilding, and therefore more threats you will inevitably be weak to. Metagames that are diverse tend to, in a competitive sense, favor Teambuilding skill. By this I mean, when two players of equal skill fight in this metagame, usually the one with the team matchup will win. That's an inherent quality of a diverse meta. This is, in part, why tournament players are baffled that newer users are able to beat experienced veterans like Bloo or McM or whoever, and why newer players like Tesung, Anti, and Ben Gay are widely considered to be the best ORAS players. They're great players, but they're FANTASTIC team builders. They are able to take advantage of the sometimes obscure but still powerful threats ORAS has to offer, while still building teams that solidly cover as many threats as possible. This type of skill is more favored in a metagame as diverse as ORAS is. To put it simply, if a metagame has threats A, B, C, D, and E, and you aren't the best team builder, you might cover threats A and B, but be unable to cover C, D, and E. However, a better team builder might be able to cover threats A, B, C, and D, while only being weak to E. This means they are less susceptible to matchup, and therefore have a greater chance of winning any given battle.
On the flip side, you can have a metagame that lacks diversity, but has only a few threats to keep track of. This diminishes matchup-based wins, and therefore teambuilding skill is not as important as in a diverse metagame. In a centralized metagame, battling skill is more heavily favored. Think about a metagame like GSC. Completely centralized. There are only a handful of Pokemon that are legitimately effective (compared to the later games), and every team carries a Snorlax and 2-3 ways of beating it. However, GSC is considered by some to be the best meta, and it is often compared to chess. In a meta like this, since there are fewer threats to cover with team building, the battling skill of each player is more important than the team building skill in determining the winner.
So those are the two different types of metagames we can have. I diverse, matchup-based one, or a centralized, battle-centric one. Ultimately, the decision as to what kind of metagame we're looking for is a simple one as soon as we answer the following question: do we inherently value teambuilding skill or battling skill more? If we value teambuilding more, then a diverse meta is what we want. If we value battling skill, a centralized meta is what we want. Obviously both skills are important, but determining which is MORE important will lead us in the direction we want the metagame to go.