I've played a couple campaigns in the beta now - 1 ME, 1 hive, and now trying a normal bio empire. Overall, I really like the change - the trade-off between using unity for hiring leaders, boosting planets, etc. vs. saving up for traditions creates interesting strategic choices, which overall give me more of a sense that I am making real choices about how my empire grows. I strongly prefer this new system to the old one based on influence - which too often fell into the "wait to have fun" trap. I fall into the camp of those who think the biggest problem is that the changes don't go far enough. I'll give a concrete example to show what I mean and then make some broader suggestions.
There seems to be consensus that unity costs are poorly balanced in early game. For instance, 200 unity to hire a second exploring scientist out of the gate doesn't feel fun. But that 200 unity also gets fairly trivial later in the game. The obvious solution, I think, is to make the cost scale with sprawl. Early game it should be no more than a couple of months worth of starting unity production for an average empire, say 60. As sprawl increases, the cost of hiring leaders should also increase - at least linearly, but perhaps even geometrically. That way, there is a real difference in play between tall and wide - tall empires, or unity focused ones can churn leaders, while wide ones can't afford to do so. It also makes thematic sense that very large and poorly governed empires find it hard to attract the best talent into government service.
More broadly, I think the biggest problem is that, for all the systems that unity is now tied into, it feels very little like it has much to do with "unity" and therefore the penalties associated with it feel somewhat arbitrary. I think this accounts for why so many players are saying they feel "punished" or that they have "less control" even though the new system objectively provides more strategic depth, not less (if players really didn't want to be forced to make trade-offs to achieve long-term goals, we wouldn't be playing a grand strategy game).
The way I see it, unity should be tied to internal cohesion, stability, and good government, while influence should be used for expansion and foreign relations. That is clearly the direction this beta points towards, and I love it. Long term, empires with insufficient unity should suffer from internal rebellions, decentralization, and eventually run the risk of splintering. This should be more of a threat the larger the empire grows geographically, much like in earth history.
To realize this, there needs to be a real consequence to running on low unity (I'll get into what I mean by "running on low unity" in a minute). Right now, the consequence is that you can't get traditions and other goodies as quickly. But that isn't fun - it's the same "wait to have fun" trap that mana based systems too often run into. Instead, a player should be free to have very few politicians, or priests, etc, and still be able to unlock traditions. BUT, they should be running the risk that negative and potentially even game-defining events will strike, similar to how crime works now.
To implement this, I think "empire sprawl" should be bifurcated into two systems. Here I'll call them "sprawl" and "overextension" but call them what you like.
Sprawl would work much the way that empire sprawl does now, giving linear increases to the cost of technologies, traditions, campaigns, etc. I would rebalance this however to make it simpler to understand and to feel less punishing. There is no good reason that the tradition cost increase should be twice the technology one. Both should increase at the same rate. So should the cost of campaigns and edicts rather than those becoming more expensive 10x faster (not to mention them ignoring the 50 admin "floor" as they currently do). The penalty should be significantly higher than it is now, maybe 3 to 4 times greater, so that it impacts play during the 200 year campaign period, rather than the diminishing returns of expansion only really hurting when you get to absurd levels of sprawl. To balance this, the cost of traditions and technologies should ALL be reduced significantly so that a medium-sized and tech/unity balanced empire unlocks traditions and researches technologies at the same rate as one does in 3.2. To the degree that less thoughtful players will find this off-putting because it seems like a "punishment" many of these changes can probably be hidden under the hood by just using an average growth formula to give "expected" costs to undiscovered traditions/techs. Playtesting can give the average rate of sprawl growth - multiply that by the time it will take to research a tech or tradition, gross it up, and you've got an expected cost. In other words, when I look at a tech option I'm considering researching, the game should display back an "expected" cost based on the notion that my sprawl will increase by the average rate before I finish the research. A tooltip could explain the calculation to the player.
Then there's overextension. This would be a separate value, taken by dividing current unity generation by empire sprawl. There should be different game effects depending on that ratio. If the ratio is very large there should be some kind of bonus to stability reflecting that my empire is cohesive and has a vibrant civil society, interdependent planets, etc. (or the gestalt equivalent). There could even be more interesting bonuses for very high unity over sprawl - like stronger defensive armies, lowered war exhaustion gain, increased trade protection and increased sublight speed in empire space. These bonuses should be useful, but need not be massive to feel good.
On the other hand, if the ratio is very small, regardless of how I'm SPENDING my unity, there should be negative effects. First, decreased stability, but that is far too easy to mitigate. Second, the bonuses for high unity/sprawl should have annoying but manageable counterparts at low unity (again, more piracy, less motivated defensive armies, higher war exhaustion gain etc.) These penalties should hurt but not be crippling (and I would not endorse getting a penalty to sublight speed, that feels far too annoying)
More interesting would be event chains that trigger when the ratio gets small. Sectors should begin exercising autonomy beyond central control. Local planets should begin pursuing their own parochial interests at the cost of the larger society. In short, I should be risking civil war. There should be ways to mitigate this, but they should cost something. For instance, a sector autonomy movement event chain could allow assigning envoys as "inquisitors" to root out disloyalty using the currently way underdeveloped espionage system. In general, a lot of the espionage mechanics could be extended to include internal control systems like secret police - allowing an empire to go on with very low unity at the cost of having to maintain a vast internal security apparatus.
Finally, to make the effects not too punishing, there should be special mechanics if civil war DOES break out, because you want players to be willing to take that risk if it means they can get the resources or obtain the size they want. A civil war could work much the way the machine uprising does now, with an event chain leading to a total war - but rather than divvying up systems at random, it should break down by sector - especially when there are sectors with lots of unhappy pops or recently conquered xenos. But unlike the machine uprising, a status quo should mean that the empire remains unitary. What I would suggest is a modified sort of total war where surrender IS an option. If the empire suffering the rebellion achieves victory, they would instantly reannex the rebels without penalty and get a significant boost to unity generation for a number of decades. If the rebels achieve victory, they get their independence as a new empire or empires. A status quo would mean that there is an uneasy return to unification. The original empire instantly reannexes the rebelling systems, but they don't get a unity generation bonus, and the systems become less productive for a decade - something akin to, but less severe than, the recently conquered penalty.
Anyway, that's my two cents. Love the direction the game is moving in, but I hope to see it much more full realized.
