First I'd like to thank TonyV for his efforts and organizing all this. It must be taking a lot of time next to whatever he usually does.
I have a few cents of my own to add to this discussion:
1) I think the best way to keep the game is if we can find a solution where the existing clients (and servers) can be used. I have a little experience in 3D modelling and the amount of time it would take to recreate the costume designer alone is staggering. Not just that, but tools like Maya and 3DS Max are very expensive and take a lot of time to learn. The open source/free alternatives have the problem of not reading/writing the standard formats (mostly Blender). The custom character design is at the core of this game, without the player, costume and power models that currently exist in the game, it will be a long time in making any replacement.
2) I worry about the legality of running private servers. I think it would otherwise be a possible solution, at least for a while, as it means we can at least keep using the current artwork/client. I'm part of a gaming community that's been running their own game servers for more than 10 years now and used to run MUDs before that. It can work if you have an organized community.
www.conclave.cc is our website.
3) As far as organizing a community, I think that for an effort like this, you need a studio with paid staff at the core. I like the Mojang/Minecraft model, where there are official releases of the game, but also a large modding community. Mojang actively talks to it's modding community and even has hired some of the most talented modders. On the other hand, Mojang closely keeps control of the "official" game, as they have seen what went wrong with InfiniMiner. They have central player database servers, but anyone can run their own world server after authenticating. I think it would make a good model for our future. Revenue would need to be based on selling updates to the engine and content to the players. Something like what Guild Wars does, but having the community run the map servers.
4) We don't live in 2004 any more, there are currently a lot of available engines for building 3D games. The top of the line stuff like the CryEngine and Unreal Engine, which could be troublesome in licensing, but I've been recently looking at the Unity 3D engine (works on my Mac) and GarageGames just released the new version of the Open Source (MIT licence) Torque 3D engine. I haven't looked at Torque 3D yet, but something like Unity 3D or any of the other engines is advanced enough and pretty easy to develop for. I don't think that the 3D engine would be the biggest stumbling block in developing a successor, there are even decent Open Source solutions. As I said in my first point, the amount of Art needed and the costs to make that (motion captured animations?), would be a much biggest stumbling block.
5) My experience in writing Client/Server games is limited to old MUDs. I don't know how hard it is to get right, but I see that some of these game engines even come with the building blocks for these kinds of things. Heck, I think I even saw a tutorial on how to build an MMO with either Torque 3D or Unity 3D. I think with a clear leader and sufficient resources, this is not unsolvable.
6) Finally I want to point at what ActiVision did years ago with Call to Power, when they stopped development on that. They licensed the source to the community (Apolyon Network), but kept full rights to the art and IP. This basically allowed the community to patch and mod the game, but without losing the IP rights. They didn't use an Open Source licence. Bringing this up in any talks with NCsoft about licensing details might be useful if we ever get that far. It would allow anyone who has the client to continue playing. I have seen efforts like this also fail in the past, just because of the costs of writing such a license, someone still has to pay the lawyers that NCsoft would need to set this up.
In short, I think there can be some leverage of Open Source, maybe through things like Torque 3D, but I think there needs to be a studio in control, with central player administration and things like that. I think maybe running the servers could be done by external hosting and local communities, like it's done for MineCraft and a lot of FPS games. My biggest worry would be the need for a lot of 3D art development, and the time and costs associated with that if we can't find a way to use the existing CoH artwork.