peter:
That shang thing is old, saw it on GX ages ago. I wonder if dee is also a copycat.
The difference between chess and RTS games is that chess has gone through many changes in its history, whereas AoE hasn't. Look at RoR, and how it caused a whole new set of strategies to emergy. What if instead of AoK and AoK:TC there had been two more expansion packs for AoE? Would it have made a difference? I think so. Every time you change the balance, it's time to re-evaluate all strategies.
For a project as complex as AoE, it takes a lot of testing to find all the bugs. Think thousands of people just to test it. If you have to pay them, and wait for it to be done before you release the game, your competition will sell a dozen games in the time that you can sell one. Asking a game company to do things this way, then, is equivalent with asking them to go bankrupt.
Having people playing the game then, and paying you for the privilege, is a more feasible way of testing a large project. Again, you get the problem that they simply aren't paid to fix the bugs. People don't buy patches, they buy new games so why make patches?
What would fix this problem, is if a company continued making money based on how much time people spend playing their game. This would encourage fixing and improving it, because if there's huge bugs left unfixed... word would have time to get around. At the same time, if the game is good, I suspect a lot more money could be made from it this way, and for less cost than simply writing new games since the marketing and basic design are already done. All you have to do is just keep your coders working on it.
I think this would solve the problem of maintaining large games.