quick note to say i pressed option 3, but would choose option 2 if went to snapshot (didn't realise you can't modify your vote on the forums!)
I'll start with what I think at the top and then see how the 3 frameworks measure up.
I think that arriving at shared mental model across the growing population of ohmies will reduce the intersections of a common ground that we can all agree with. I've agree with those who've come before me that one way of looking at Olympus is as an operating system (OS). Within this model then, the OS's which I admire most, and which are most widespread throughout the internet infrastructure we all rely upon (linux, devuan, early debian, freebsd) the governance for what makes into the kernel is, or historically has been (for the most part), extremely conservative. With these two perspectives in mind, my personal opinion is that strategies which decrease surface area, which do one thing and do it well is a sound approach.
OTOH i have a radical opinion that optimizing for minimizing the chance or avenues for siphoning funds, preventing free riders etc are optimizations and energy usage along a lower order priority than racing against deflating value of DAO owned OHM (i expanded on that a little in OIP-49). I understand this is likely a contentious take - but I would personally rather see ohmies paid extremely well for any work they do in the DAO, and see experimentation happy with an openness to failure, with less of a restrictive focus on ROI. That's not to say a free for all - that's to say that with the humans we have congregated around Olympus, with the systems which we already have in place, that any loss from free riders, or siphoning would be minimal compared to expanded bandwidth of not trying to implement systems which come from a scarcity framework. I would rather 99% of dedicated ohmies be free'd up to 100% focus on value creation and making bold advances, then worry about small change loss. Dilution of DAO owned OHM far outweighs the pace of any bad actors could survive in a DAO which is humming along in an increasingly coordinated fashion.
With all that being said, and back down in reality. I think in reality that a conservative approach with the DAO owned OHM/ treasury is probably more pragmatic/practical with where people are coming from. I think we should aim to move towards permissionless everything. i think the future is permissionless curation. so something like rari fuse pools, but where curators build a rep for assembling pools of recipients and then people can redirect yields to the curators/pools which align with their priorities. i imagine if a different set of worldviews had informed the selection of a different set of charity donations then a different segment of ohmies would have put up a level of resistance. historically these governance discussions/debates have been nessecary but we now have permissionless technicques and tools which allow the grassroots emergence of a plurality of priorities to emerge, and in our case with olympus as a base stem cell operating system.
Specific feedback about the frameworks
My feeling about framework 1 is that it's a response to the threat that people would be implicated in decisions/value flows which they don't agree with and don't have a meaningful way of opting out of. I feel like framework 1 would leave another group of people feeling like there's a decision that, if it went through, they don't have a meaningful way of opting out of.
My feeling is framework 2 is the best of the three options. I think having deliberation and a slower process and requiring a higher level of buy in (number of unique wallets voting) is a good precedent.
My feeling on framework 3 - i would rather have seen this be something like a pause / moratorium with some process for finding where the common ground is - if any, with a decision that he widest number of people could agree with.
All this can be summarized as permissionless tooling FTW