I agree with Deck Knight in pinning Concept Assessment as the stage with the biggest failure and threat discussion as the second biggest. Korski basically most of I would have wanted to (and he even pointed it out to me), while the rest I already said in the latest Smog issue. When people were saying "no counters", we should not have stopped there and we should have identified strong, reliable checks. However, I stand by my intention not to make this project about luck, because when people went into the playtest, I didn't want people to get angry because of hax, but rather, I wanted people to be forced to blame themselves for the consequences of their own decisions. I found that a much more interesting prospect than a study on hax.
I do have an idea for every stage, namely:
For typing, I initially had a very different idea from what actually happened. I felt that a typing with immunities (both offensively and defensively), priority weaknesses, and some safe aspects such as resistance
to Stealth Rock would have been a good way to "ground" the project. I felt that a Pokemon that you could switch in over and over again, but whose decisions after switching in mattered every time, would have been more manageable and focused from a process standpoint than a typing with big weaknesses to U-turn and Stealth Rock whose aim was to go all in. A few weeks ago on IRC, I and a few other people massaged this idea to come up with Normal / Ground, Flygon stats, Guts, with Wild Charge as its only significant non-STAB coverage move (well, maybe less Attack because STAB Guts Facade is SCARY). I'm not convinced by bugmaniacbob's assessment that something like Guts or Flare Boost would have been boring because it doesn't really exist in OU and we'd be able to get a better idea of how it could thrive.
For the ability stage, as I said in The Smog, I think that all of the abilities had good justification, but only one should have won out. Weak Armour was great in principle (I personally disliked how I had to consider high physical bulk because of it, but other than that...). Illusion was weird, but I still think it made for a good experiment. I still think No Guard might have been cool by itself, turning threats like Politoed and Terrakion into reliable checks (I never really cared about inaccurate coverage moves). I partially blame myself for focusing so much on the principle of it instead of focusing on making reliable checks.
Stats were weird because I (and I'm sure others) felt so much pressure from so many differing views. In particular, I did not care about mixed attacking because being checked by Chansey is not exactly a big deal, yet I felt I had to up the Attack to get votes/slated. I also never really considered boosting moves; I didn't really want boosting moves at the time, and I didn't really expect them to be a big point of contention. If I were to redo the stats without any pressures, I'd probably go with something like 110 / 50 / 99 / 117 / 30 / 94.
As for movepool, I think that the coverage moves were excessive on most of the submissions. Out of the significant coverage moves, any one
of them would have been sufficient. I probably should have voted for jas's movepool in the last poll, heh... Not to toot my own horn, but I still like my submission
Also I love Rhys DeAnno's assessment of Gengar, which managed to discuss Substitute of all things in a way that never really happened during CAP 4's discussions.