I corrected my list appropriately with Birkal's update about half an hour ago, but lando-t has just enough usage in one of the two percentages to break the 5% barrier. (Compare to keldeo, who has just enough usage in the OTHER percentage.) I chose to cut them both and put them on the list for consistency's sake, and since they were also above 5% in the 3-month stats, it felt logical to me. Basically I was looking at them thinking "okay can I justify cutting one but not the other?" and it was 2 percentages out of 3 being over 5%, so the easiest thing was to take them both.
Here are the Sept-only stats right at the bottom of the barrel (left being percent usage, right being percent real usage):
| 45 | Keldeo | 31591 | 5.097% | 24245 | 4.823% |
| 46 | Landorus-Therian | 30597 | 4.936% | 25693 | 5.111% |
| 47 | Metagross | 30310 | 4.890% | 24557 | 4.885% |
And here are the 3-month stats for those same mons:
| 45 | Keldeo | 5.183% |
| 46 | Landorus-Therian | 5.087% |
| 47 | Metagross | 5.025% |
Metagross is legal now though, definitely.