Disclaimer: i added some anecdotes/comments from what i can remember of the workshop. my memory is far from perfect.
Simply put, DeFi wants to evolve or move beyond the veCRV model.
The general consensus is that, given the right incentives, a DAO can be bribed or simply incentivized into doing just about anything, including harming the long-term viability of another DAO that it aggregates, and even self-annihilating. The main reason for this stems from the externalities that are ever-present to any DeFi protocol, simply due to its existence in a larger, highly complex ecosystem.
It was also largely agreed upon that bribe markets will find a way to exist, no matter what.
The mechanic of locking tokens in exchange for governance power is stronger for longer locks, and allows governors to commit themselves to the system more deeply. However, there were multiple comments on a more controversial statement that challenged the usefulness of increasingly long locks.
The group’s answers were evenly split on whether or not lock lengths longer than 1 year had diminishing returns for the protocol, but comments given during the larger discussion pointed to a more nuanced take, where the curve of the lock-time vs power graph could be altered to achieve better results. Perhaps an exponential curve would be better than a linear one?
Lockers have their limits on how long they are willing to commit.