checkmater75, I'm not giving you (or Hashtag) any particular attitude. Head Smash Scrafty is evidence that I'll listen to your set when it is logical. That's a move I had never even considered on numerous teams on which I've used Scrafty, but it's a good choice. That Raichu set however would get an instantaneous QC reject 1/2 from me. And from the looks of it, it would get one from Memoric too. I don't care if it works on one team. It would never be a recommended analysis set. Fake Out, Encore, and Feint together is so duplicative I can't even. And beyond that most variants of that set are still largely outclassed by Infernape, who learns all 3 moves and offers a real offensive presence (or at least bluffs one while wasting 3 slots on roughly the same effect).
So why don't I approve checkmater and then just make a dozen changes? Because analysis writing is not a system where the writers come in and say "I'll write this mon - QC tell me what the acceptable set is." The set is expected to come in quite close to correct. It is not the job of QC to radically fix things. If vast changes would be necessary, the analysis should be rejected. Most of QC's effort should be spent on making sure the actual ANALYSIS part of the analysis is good. So if I believe I already know the set he'd post (since he's posted it before...) and I know that it is not up to snuff, I'm both saving everyone time and maintaining quality in the dex by saying no. And while I could maaaybe support a Raichu analysis, it is such a borderline mon already that I'm not rushing to approve it for approval's sake.