"Fun" is not really what I personally hope to achieve, although it is still important. Apparently you don't agree that the game should be fun?
We want a patch where all civs are viable. Arguably, we have achieved this--or mostly achieved it--by now. From here, the goal should be to broaden the versatility of civs, slowly, and not drastically. This means that most civs should at least have some options on TP maps, non-TP maps, and on water. Obviously perfect balance here is impossible, but we can do better:
I know we can. Ideally, every civ should at least be
playable in the lategame as well, as this makes win conditions more reliant on skill, rather than civilization, although it is important that civs still differ in strength at different stages of the game.
Ideally, the end product requires interactions between players throughout all stages of the game. I see people complain all the time about how AoE3 is often about massing units for 1 big fight, and whoever wins the fight wins the game. How do you avoid this? Making scouting and adaptation more important is a good start, and you cannot make that important without providing room for civs to adapt to different situations. Perhaps making win conditions less obvious, but provide more of them: you want to stack your win conditions in order to win a game. How do you accomplish this?
Take a matchup like Aztec vs China, for example. In this matchup, it's completely unwinnable in age 2 for China, and equally unwinnable for Aztec if China reaches age 3 with their economy and infrastructure mostly intact. Aztec's win condition is obvious: prevent China from reaching Fortress, or cripple them so hard that they cannot do anything once they get there. China's win condition--obviously--is to avoid this. This makes the matchup so one-dimensional that it's not even worth scouting by either player, so you simply rush as hard as possible as Aztec, or FF/turtle as much as possible as China. There are no adaptations to be made, nothing to really think about. To fix this, you have to make China's Colonial a little bit better (in this case, it's easy, just make sure that Qiang Pikemen actually counter coyote runners to the same capacity that they counter cavalry, which at the moment they do not), and Aztec's Fortress follow ups better, but you maintain some of China's weaknesses in Colonial and some of Aztec's weaknesses in Fortress.
With this done, it becomes much more important to know what your opponent is doing. Suddenly, it matters to Aztec whether or not China is staying age 2 because it determines how fast they need to play the matchup.
Is China FFing? Should I delay 5v to get more units out earlier while he's aging? Is he staying age 2? Should I do a slightly slower build to get more units out during the times he chooses to age with this semi-FF?
Equally, as Aztec's Fortress options are better, China needs to consider that staying age 2 could result in getting punished by an early Fortress timing or something, and need to actually scout to see what Aztec's plan is.
Suddenly, Aztec has multiple win conditions: (1) Abuse China's Colonial weaknesses as much as possible before following them into the Fortress Age, and (2) Either survive China's Fortress power spikes or outplay them on the map in some capacity. At which point the next stage of the game also becomes relevant. Before, it would be okay: Did he take significant damage from my rush? Yes (China resigns) or No (Aztec resigns).
And there are lots of matchups like this. Even
mirrors would benefit from these kinds of changes, as there's no longer a clear "best build" in all situations that you must adhere to. In a Brit mirror, for example, maybe you should be concerned that your opponent could actually go for some sneaky semi-FF or something, and have to actually check out his build, instead of trusting that muskets + map control is going to win you the game against anything.
At the end of the day this will make the game more about skill and game knowledge and less about massing up with the same build order every game and hoping that you can win that big fight this time around. And, yes, hopefully the game will be more
fun as a result.
I don't think any of this is unachievable. Perfection? Unattainable, certainly. But doing better than we currently have? Absolutely.
And so I see no reason to stop here. If it means treading more carefully with our future changes than in the past, then so be it: we can slow to a walk instead of a run--a crawl, if we must. If it means that we stop and take the time to redo some unpopular and arguably unsuccessful changes, then that's great too, but if you want to reach that point then you need to be less pessimisticly hateful and more optimistically open-minded and patient.