I'd like to see more discussion on the 2 closest suspects in Smogon's history: Aegislash and Genesect.
I'd like to first note Genesect because it feels like if something in Ubers is going to drop, it's going to be Gene. It used to be Greninja's big brother when it was around in early XY, but now Gren way surpassed it in terms of power and versatility, and while it was easily Ubers, it wasn't absolutely meta breaking like most Ubers (ex Mence). Especially with the light power creep introduced by ORAS and seeing how the ORAS meta's capability to at least somewhat handle a mon that was basically LO and Scarf Gene with better coverage rolled up into 1 set, I really don't see Gene's power, versatility, or access to Download being as defining as they were in XY. Especially not on only base 99 Spe.
Aegis on the other hand is sitting on a slightly higher pedestal, unlike Gene it has the full ability to warp the metagame around it, being the deciding factor on what is viable and what is not. However, in Aegis' case... that's a good thing.
This is where we start digging deeper. There is one really notable problem right now: Lando-T. It's usage has far surpassed anything that any OU meta has ever seen, its prevalence is so all-encompassing that at their highest usage when they were together in a metagame, it currently has more usage than the combined usage of both Aegislash and Genesect. Why is this happening? The answer is pretty obvious: there simply aren't other good pivots in the tier. If you need glue, it's Lando-T. It's the only thing left that can answer the sheer volume of threats ORAS has, and it effectively prevents many things from tearing the tier apart. This is a serious problem. Much larger than MGross, much larger than Sableye, much larger than Shadow Tag. The game is currently stuck in a fixated state where if you're not running stall, you're nearly forced to run Lando-T because you don't have other options. The threats are too large in number and are too strong in XY. This is a problem that the game's been having since Gen 5, the questions too far outnumber the answers, and the threats are much stronger than the Pokemon used to check them. This causes a game heavily biased by team match-ups, where your opponent could be carrying something you simply do not have the room to check, causing a near instant loss. Lando-T is used on so many teams because it shrinks the number of instant-loss mons by an incredibly large number. It is the last good pivot left in the game.
This is where we go back to Gene and Aegis. These mons are threats in their own right, they both are versatile, hit like a truck, and have overall fantastic stat allocation and typing. This is how they got in Ubers, but maybe we should've looked deeper into the meta shifts before we did this. Genesect's ban might've caused the worst meta in Smogon's history, it took multiple bans that happened over the course of months to finally put the game into a state where HO teams weren't ruining the game, and this fault game from the fact Gene was the glue that prevented all that from happening. Gene no longer has this power, as the types of teams it negated are no longer viable due to these bans, but it still is a strong asset to checking offensive threats in the meta. Aegis also dictated the game, making many otherwise OU-capable mons get stuck into BL or UU because they simply can't cope with it. Both these held the game in a state where dozens of Pokemon were locked away, unviable as threats because these Pokemon existed. They were the perfect pivots. They did what Lando-T is doing now, but also doubled as very good mons on their own. Ultimately, this is what the game needs. We currently are stuck in a meta where it is no longer possible to build non-stall teams that that sufficiently check the metagame (and it's only possible for stall to do that because of M-Sable), and this is where we need to look at suspects not in the perspective of their individual ability, but the effect they have on the game. Many people wanted to ban Aegis initially because it would bring more versatility into the game, but I believe after seeing how the Gen 6 meta has progressed it's pretty clear that versatility isn't always a good thing. Sometimes we need these incredibly strong pivots to prevent many threats from being viable in OU, which allows a meta that is less dictated on team match-up and more on teambuilding and player skill. I feel like both these mons would help weaken the power of top threats such as Mgross while not being completely overbearing in themselves.
As a final note, I'd like to discuss MSableye in regards to the potential of these mons dropping. As is stands, with Sable banned, stall is done. We will actually be left with 0 playstyles left in the game where the game comes down to mostly player skill rather than team match-up, since no other playstyles are currently capable of avoiding bad match-ups. Even Goth doesn't dictate an instant loss for stall due to the strength of MSable stall teams. I think this change is incredibly bad for the game if not complimented with something else, and that something else is Aegislash. If Aegis were to drop, I'd fully support a suspect test for Sableye, since now stall would have the wiggle room to play around with many different metas and team building styles, rather than being stuck in a place where they're just building around Sableye. Aegis not only is a strong asset for stall teams that would prevent the forced usage of Sableye on stall, but its presence in the meta pushes out threats that Stall currently has to worry about, many of which are Sableye checks. With Aegis in the meta I'm very confident that Sabeye's presence would no longer be needed, and if Aegis and Sable were swapped out we'd end with stall having less restrictive teambuilding while building for and against stall.
tl;dr- Aegis and Gene have a positive effect on the game because they encourage a game based on teambuilding and player skill rather than team match-ups. We should suspect test to drop them. Sableye as it currently stands is fine, but if Aegis drops it needs a test.