Regarding unbonding queue mentioned here a few times:
I have recently had a look at the PR, and admittedly my estimates of it being “ready” was incorrect. To properly implement this feature and have a high confidence that it is solid, we would need to spend (based on my estimate) at least another month of heavy development (followed by overseeing the deployment and all the logistics of that) by our staking experts, which will only slow down delivering all of the above.
The junction I see is then, assuming we move forward with implementing the above, is as follows:
- We investigate if we can break this down into phases, and foremost implement the subset that will make nominators’ APY different than validators, their slashing risk is removed, and their unbonding time is 1 era. This means in ~Q2 next year already, nominator staking will drastically be different and unbonding queue is not needed
- If we identify that implementing the above is entangled with the full scope, and it takes e.g. 1 year to fully implement it, then there is a stronger argument to push unbonding queue forward in parallel, as it will serve nominators for a longer period of time.
Personally, I don’t think faster unbonding time for validators is particularly important, and it is in-line with making validators as aligned with the protocol as possible. Needing to wait 28 days without any rewards will make a validators more convicted to remain as a Polkadot validators.