This is a great example on how misinformation usually spreads. Bear with me here, and I'll give you an insight into the dynamics of user support, forum boards and what's it like dealing with components outside of your control.
In the case of SleepyC, as far as I know from talking with Andrew, the problem appears with one particular motor (8108 I think) with heavier cameras. 6208 motor works, as does a NEX with the 8108. So there's one combo giving troubles (GH4 with 8108). So that's problem #1.
Then, in the case of Scott, the setup is totally different in terms of motors, camera and gimbal; even the problem is different! his yaw won't even tune (the CP detects bad motion on the yaw and drops into 2 axis mode). He tried different CP and different IMU, and the problem persists, meaning it's 99% something with the mechanics or motor and it's just picked up by the CP (as it should!). We have extremely detailed logs (as you all know) and we can see the CP pushing power up on the axis trying to get some kind of clean movement but that doesn't happen. This analysis is extremely precise, up to us telling users they have a cold solder joint and such. So that's problem #2.
Next, someone will chime in with a Cinestar gimbal with rotating landing legs, which mechanically compromise the gimbal (ever wondered why DJI never went with that, despite the added complexity of retracts?) and again "CP yaw doesn't work!". Problem #3.
Then, someone else will have a too soft of a damping setup. He'd complain about yaw too (and he should! because when rotating the gimbal, the same exact force is exerted back on the dampers - it's kindergarten physics; too soft of a damping system, and the yaw doesn't have anything to work against, and the dampers will just deform instead of the gimbal creating a clean motion).
Problem #4.
Then, some guy up in Canada buys the CP for his $129 GoPro gimbal, which hangs the CP on the roll arm. Being that it's a GoPro gimbal, it's super light compared to the mass of the electronics and the wires (which are completely negligible in something the size of a MoVI). He has a battery coming down from the top and it has a quite rigid cable. Heck, even the stiffness of the USB cable interferes with the yaw tuning. So the setup is really bad and doesn't allow the yaw to operate freely (and even if it did, heck it's the cheapest design and materials you can get). Then he comes and complains about his yaw problems. CP doesn't work well on yaw! Problem #5.
Another guy then uses a great, well sorted gimbal. It's even handheld so there's no dampers to mess up anything. However, he doesn't tie down his gimbal properly for the autotuning. It seems to be held fine, but during autotune the gimbal shifts and slides a bit in place. How bad can it be, right? Well, it completely messes up the autotune process (which exercises the mechanical structure of the gimbal, and for that it needs the gimbal clamped in place). The pitch has the mass of the rest of the gimbal to lean against, and the roll at least as the yaw (and battery) to lean against, but the yaw is completely free to wiggle around. Yaw tunes bad, and the CP YAW DOESN'T WORK! Problem #6.
There you have it, SIX people with yaw problems! They'll all convene here and say "CP YAW DOESN'T WORK". Soon enough the trolls from RCG pick that up and repeat the sentiment that CP yaw doesn't work. And if we say we can maybe improve on some of this stuff in software, it'll be used as an automatic admission of guilt that the problem is in the CP.
So six completely unrelated problems, with the only common denominator being the CP - although it would be ridiculous to claim the yaw doesn't work on the CP, would it (being that, well, 99% of CP users have three axis systems). But it would still reduce the discussion into CP YAW DOESN'T WORK.
In reality, there's only one problem worth investigating here - and that's the first problem with the 8108 motor. The others are completely outside the controller's domain. We can't do much about it. Yes, you can take an Alexmos and tune it badly enough so it works in all cases. It'll be all over the place, and you wouldn't want that from us (or else you wouldn't have bought our three times more expensive hardware).
Before you dismiss the role of mechanical setup, consider that between the first revision of the F1 and the current third revision (it's completely different than the photos we published right now), the autotune gain numbers went up by 3x, as did the usable focal length (!!!). The only changes were mechanical hard parts. Same motors, same controller (running 1330 software you all use) between all versions. The only change is the mechanical design of the gimbal - and there's even more room for improvement on that.
