It seems there is an assumption here that each community (really sub-community) is considered to have equal influence and impact to the larger Eden community co-existing on the same platform (i.e. sharing inflation and making governance decisions impacting their host blockchain). The idea that they share equal amounts of inflation makes that very obvious.
But I think even the requirement to reach unanimous agreement on blockchain changes subtly relies on that assumption. Just consider the following thought experiment. Assume that there are already two sub-communities (let’s call them A and B) of relatively equal influence and number of members that already exist. Both are sufficiently large to meet the threshold for existing as a sub-community (let’s say they both have 10,000 members). Now 1,000 members of sub-community B get the bright idea to secede from B and form a new sub-community C. This gives them more influence over governance decisions on the blockchain (not to mention a considerably larger share of the inflation funds than what they controlled acting within sub-community B). Now in negotiations on blockchain governance, sub-community C representatives can threaten to hold back progress on governance unless concessions are made in their favor. Those 1,000 members making up sub-community C now effectively have the same leverage to get compromises in their favor as the 9,000 members remaining in sub-community B. But if they had stayed they would have been forced to compromise with the other members that are 9 times their size, and because of the size disparity they are significantly less likely to have gotten their interests represented by whoever the representative is coming out of the Hierarchical Governance (HiGov) election process from sub-community B. It is clearly to their advantage to split into a new sub-community if they can in order to get more influence in the blockchain governance process. The incentives disappear when they have all split enough that all remaining sub-communities are of roughly equal size not much greater than the minimum threshold to form a sub-community.
In addition, doesn’t the self-selection afforded to members by allowing them to easily create their own sub-communities counteract the benefits of randomly selecting from large groups to tap into the wisdom of the crowds and to counteract the ease of collusion, regulatory capture, formation of party systems and the like?
That said, obviously the right of a member to secede from a sub-community they do not like is paramount. Though I believe that they should post a sufficiently large bond to join a community and though they can leave at any time, their bond should remained locked for some time so that there is sufficient capital to seize if a dispute resolution process finds that they violated community rules shortly before seceding but the process was only able to get to that ruling some time after they seceded. This ensures they actually do have skin in the game while they remain a community member given the practical realities of how long it takes to actual go through a dispute resolution process and get a final judgement.
Of less, but still significant, importance is for them to be allowed to form their own new sub-community that can participate in the greater community governance process (for example the one that decides on blockchain governance decisions or on how to split the available stream of funds among sub-communities). But the rules for forming such a sub-community could be strict and furthermore I don’t believe it makes sense to just give them privileges of equal influence to other more significant (by whatever metrics one cares about) existing sub-communities. It is not clear to me though what the rules governing that should be to mitigate against abuse and avoid creating perverse incentivizes like the above.
I would like to see more exploration into how governance could work at the level of these sovereign sub-communities (maybe there is something analogous to the international relations among sovereign states?). In particular, what would be a more reasonable way of coming to consensus on the relative importance and influence between sub-communities of different size and how inflation funds should be fairly divided among them to avoid incentivizing incessant splitting of sub-communities until it reaches a flat structure of an untenably large number of sub-communities which can each unilaterally hold back any progress of blockchain governance with a single veto (the probability of that gets very close to one when the number of actors that can unilaterally veto progress gets to a very large number).
Perhaps one approach may be to instead apply HiGov among the representatives of each of those sub-communities to determine a smaller number of leaders that handle blockchain governance. But then I think this is just a degenerate case that is effectively equivalent to enforcing only a single community except that people are allowed to severely constrain the random selection/grouping aspect within the first three levels of the HiGov process via self-selection into sub-communities which I think would somewhat counter many of the benefits of randomness in HiGov progress mentioned earlier.