First, I’d like to note that this is my personal opinion and not L2BEAT’s governance team.
I’ve discussed this idea with Event Horizon on a number of occasions, and below is a summary of my thoughts on the matter, which I’d like to add to the discussion.
First of all, I like all experimentation with governance systems. I think one of the best things about crypto governance is that we can experiment with decision-making at a meaningful scale and with meaningful stake.
At the same time, I also have a strong opinion that voting itself is not the essence of this decision-making, but only one step of it, mostly optional. It is a much broader process consisting of idea generation, spreading awareness, stimulating debate, building consensus, etc. Voting itself in crypto is quite simple and frictionless, my personal thesis is that governance problems stem from deficiencies in these other aspects, not from so-called “voter apathy” itself.
That’s why I don’t think that simply “giving a voice to retail holders en masse” in itself is going to positively impact governance in a meaningful way. I’am fully open to be proven wrong though. 
Having said that, I think that what Event Horizon is proposing could be a good foundation on which to build further governance experiments.
For example I can imagine this pool to be used only for voting early as a signal. 5M votes is not big enough to swing most votes, but it would be enough to signal support if those votes were cast early, for example in the first 48 hours after the proposal is published (and if some quorum is not reached in those 48hrs the pool abstains and the opportunity is lost) . It incentivizes the proposer to raise awareness of people participating in that pool early on, at the same time people participating in the pool have incentive to pay attention to what’s happening cause if they don’t vote early they can’t vote at all.
Another example is a very interesting concept already utilized in Starknet as a Builders Council (Council’_council). This is a pool with huge voting power given to the elected Builders representatives from Starknet ecosystem. I personally treat it as a kind of a Builders Union that can effectively veto governance decisions if strongly disagress with it. Similar concept is being explored by Optimism with their Anticapture Commision (Anticapture Commission - Metagovernance - Optimism Collective).
Obviously what I have given above are just examples of what is possible. I’m not suggesting that this particular proposal should follow any of these routes, I just wanted to demonstrate that this is a valid experiment that has been tried in one form or another in other ecosystems, so we should not be afraid of it, but rather try to experiment with it ourselves.
What’s most important to me (and what I think is currently missing from this proposal) are mechanisms that will accompany this voting pool to ensure that the community behind it is actively and meaningfully involved in governance. But I think we can safely start by agreeing to set up the pool and let Event Horizon expand and experiment with these mechanics later.
On the technical side I have three suggestions:
- I don’t fully understand what’s in it for Event Horizon, there is no mention of any compensation/reward for providing this mechanism. I understand that the mechanism/protocol itself is supposed to be public good and is provided for free, but I want to make sure that there is an incentive for Event Horizon to keep working on it and experimenting with it on Arbitrum, so that it creates value for the DAO and does not just exist there.
- While I’m not against larger pool size, I don’t think it’s necessary. This should not be a real competition to the voting power that comes from owning and delegating tokens. I want to avoid controversies where this voting power coming from the DAO treasury swings some important vote by a small margin (imagine such DAO-sourced voting power deciding on an issue such as the Wormhole vs LayerZero vote in Uniswap last year).
- I would prefer that this voting power not be optimistically extended beyond an initial period, at least not in the beginning. Rather, we should have a vote in a year from now if this experiment was meaningful.