To revive this discussion...
It came to my mind, that we are/were talking about long term playability issues, such as, that people would get bored too soon, etc.
Now, actually, we never tested AWS' long term playability.... So I'd like to make a little suggestion:
Why don't you, Sami, consider to run a "long-game-test" with 20 minute days (would speed up real time vs. game time ratio a lot), and actually make a game world from 1950-2050, at normal cost rate, or maybe, considering the faster time-elapse, at a slightly higher cost, to cover the server-load and running cost.
One could let number of players rise over time, to also allow more competition once more demand opens up ingame.
We could then collect player responses, and thoughts, without having to come up with new features, that would eventually be hard to program, and see, if it works out without these. Of course they can be added in later versions, but just to see if the game actually works for the long times already.
As an Idea for the billing: make it 10cr to join, and then 7cr/real life month (versus 4.2/mth) to make up for the higher workload, and also to finance the extra capacity, that may be needed.
cheers for considering,
P.S. with the increased pace, the world would probably not even be as long as one and a half real years... actually: 508 real days, including the 24hr start. -just asides