The usage system already accommodates this. As the number of mons grow so does the number of tiers. That's why RBY and GSC only have 2 tiers. If the player base supports it, then we'll add official usage tiers as the number of mons grows. In addition, since Game Freak power creeps, mons that are OU now can and likely will be pushed down to lower tiers over time. Worst comes to worst, we'll just have a huge number of mons in NU; even in NU a meta establishes such that not every NU mon gets used.
It takes us 1-2 months to determine if 1 mon is broken in OU. Think about how long it'll take to do it for 12-15 mons. That is at least another year on top of OU suspects, perhaps 2, perhaps more if you test mons that were banned, and you yourself can't determine a good cut-off percentage for OUdown because the "broken mons" are scattered throughout the usage stats. Further, you can't just ban 10 mons and call it a day; need to ban about 25 mons to create a tier between OU and UU. As I said before, ban too few and it's just OU with salty bans, and ban too many and it's just UU with even more work required to balance the tier. Also, how are you going to find 25-30 broken mons to ban? I'm more than willing to bet you're going to start banning stuff that isn't broken to meet that quota. Come testing time, when something in OUdown is deemed broken, are you going to ban things that would have been checked with mons you banned or ban the mon in question? If a random check you brought in starts upsetting the tier, are you going to reban it or further bring in banned mons?
Your proposal has a huge number of logistical problems, and what's your net result? A tier that's about 25 mons smaller than OU. A player base made entirely from salty OU players is not nearly large enough to merit that much work. If you think that OU has too many problems, your best bet is to just participate in suspect tests, get reqs, and cast votes.
