I suppose I phrased that wrong. The way I see it, people genuinely are not putting enough effort into how they vote. Reaching the vaunted 1655/65 is not too terribly difficult, and from there, they once again have to justify their vote. I know that can work to weed out voters that don't put effort into their votes, but in the end, we're putting a band-aid on a wound that obviously needs stitches. We can phase out every shitty vote, but the final aim should not be "throw out shitty votes" but "make the community less likely to cast shitty votes".
Remember where I noted that the side effect of the metagame as is would be that two main lines of thought would arise: OU will say that Latias is easily managed because there are so many counter, while Uber will say that Latias is on every team and counters are overly necessary. Jumpman's answer was to simply say, "These votes suck, why do we care about them? We'll just throw them out." This is the wrong way of going about fixing the problem. The safeguards in place don't address the problem of a faulty metagame.
At this point, I am only concerned about one thing: "How can we refine the Suspect Test ladder in order to produce more useful results, both for statistics and experience?" I believe that the suspect should be dropped into the Standard metagame, and that a separate Suspect Test metagame should exist comprised of the metagame before the Suspect. This will have the effect of correctly gauging the effect of the suspect on the metagame, rather than creating a new metagame solely for the purpose of using said suspect. This will force most players to look at the effect the suspect has on the metagame. Alternately, many players, after playing in a metagame with the suspect enables, will want to return to the old metagame in order to compare the two, which is where the Suspect Test metagame comes in. People already have tons of experience in the current metagame, so the Suspect Test ladder will act as somewhat of a refresher course on the pre-suspect metagame. The end result is that you get a more informed public, more useful usage statistics, and a Suspect Test that works.