Claaarky, it seems somewhat odd to me, that you've jumped to these conclusions with zero recovery. I understand your theory and like many other people, it's something I've explored. There maybe 'some' truth in it.
I also agree, that many will know bits and pieces about Panda and hold on to it. I have several findings from my own testing that I don't share, simply because I don't want Google to change things again. But I don't believe anybody has 'cracked the Panda code'.
But as far as exit rate goes, it maybe factored in to Panda. But it's certainly not the missing link.
Panda is a quality score and any one thing can tip the balance either way. So if exit rate is a factor, you maybe able to swing the balance by improving it. But my guess would be you'd need to fix a lot more to swing the balance. It's not all about fixing things either, it's about earning trust and proving your value.
I'm not dismissing your claims, I just don't want them to be blown out of proportion.