I had suggested Tantathia for a couple of reasons. First, it's not a kingdom being run by any involved person. Unfortunately, Jason seems to have completely removed himself from the group, but that does mean no one player has extra weight on what is "true" about the region. Second, Tanathia has had a lot of turmoil in the past in-game decade or two: the Great War, the division by Barloz and Aruthien, the invasion of the Kami, the death/rebirth of the Fire King and Ice Queen, the destruction and recreation of Mount Fum. Essentially, the map we have in the wiki no longer has any relation to the countryside that may exist today. This is perfect for this kind of game. We can add/remove whatever we want; we have ready-made ruins to explore; some history we can draw on, but only if we want to; and the rest is wild and uninhabited. Lastly, it's an area we don't use much, but is pretty strategically located between some major regional powers, and doesn't have an established system of government that we'd have to check in with.
My next suggestion was to make all of the PCs relevant to the outpost we are establishing. No PC can have "adventurer" as their day job. This dovetails very nicely with 5e's backgrounds. It gives a good reason why your bard is a good cook, or your fighter is a smith. Normally, these backgrounds and skills are irrelevant in normal gaming, and players are nudged towards picking backgrounds with adventure-friendly skills/proficiencies. Further, unless your PC is holding weight in the non-adventuring season, you are going to be hard pressed to explain why the rest of this town keeps them around. Sure, we'll likely need some NPCs to do some things, but I'd like to see the PCs involved in the town as much as possible.
I like Greg's plan of having at least 2 PCs per player. I would suggest that they would need to be fairly different from each other as well. Picking a fighter with a big sword and a ranger also with a big sword, barring some very distinct other choices, basically just means you are playing the same character with two different names. I'm not suggesting that you can't do this, but this is an opportunity to explore character options. Relatedly, I'd like to set out early that "retiring" characters should be expected. If you play a character for a while, and don't like it, retire them and roll up a new PC and come up with a reason why they join the village. Additionally, PCs are going to die and, while this often sucks, I think it's necessary for the game. Heroic sacrifices are something we've done in games, but "the goblin rolled a crit and you failed your death saves" isn't something we tend to allow. But, if the expectation is that PCs are semi-expendable, we can start allowing this from time to time. It would be dependent on the PCs, players and GM to decide if a character is worth raising, rather than determining if it's okay that they die.
Finally, I don't want anyone to feel compelled to run a game if they don't want to, nor do I want anyone to not run a game if they do. This was something of an issue in the GC. Some players seemed more hesitant to run sessions, and I think they felt forced into it (especially at high levels). Whereas, I know I wanted to run more games than I ended up doing, and I suspect Greg feels the same. With Lee potentially joining up, I wouldn't want to make him feel compelled to run any games either, especially as he's still working out the nuances of all the rules.
I think we should open up the discussion about what sorts of PCs we'd each like to bring (I'll probably be bringing my Warlock that I built as a backup for Strahd), and what sort of town we'd like to build. For those not actively running PCs in adventures, I think it'd be great if you were involved in the activity of the town - who is there, what they are like, what sorts of things come and go, etc. I'm interested to see where this concept goes, and interested to see what everyone else's input is.
Pillows are designed for relaxation. If they are fighting, what hope do we have?