Indeed. The goal is to give pure soloists not belonging to a group the opportunity to experience evolution, or even the opportunity to players of low ranking teams to experience high evolution. jeffs's initial proposal starts from the fact that there is quite some inequity in the present evolvers system (To be an evolver, you must join a team. To be a top evolver, better you join a top team !).
Following your suggestions, I see 2 types of "services":
1-OR a group for all beginners who want to share their solution with any other player, in order to learn (arrangement can be made in global)
2-OR (another) group for top soloists and evolvers who want further improving a top soloist solution at end game.
Both group have the same kind of rule based on freedom, openness, neutrality (the administrator should be a non playing, non logging in, clone), self-discipline to leave the group asap etc.
I do not see any problem with group 1. Like there exists a global chat, this would be a global group.
For group 2, I do not see more risk of reducing the number of clusters than in current life with open group and public recipes. Group 2 could reduce the inequity mentioned above:Any pure soloist or any player from low ranking team can get some high evolvers experience visiting this group for a short period. But group 2 has a high potential of becoming top group if many top players share there. Would it endanger the concept of closed top teams (and "closed" lonesome soloists)?
Remark:
The sudden appearing of Soloists virtual group in top ranking for several puzzles like 880 suggests that this idea might be feasible without new dev. I found this appearing very funny (thanks johnmitch ! A new difficulty for the real teams, because the sum of best soloists has a high potential I think. The Soloists virtual group concept does not endanger the cluster problematic, but it does not solve our current question. A similar group "with share allowed" is thus needed, and feasible.
Knowing that no share can be transported from one group to the other, I think the current system is safe enough in order to avoid diminishing the desirable number of clusters.