Woa the Doomsday Generator is quite OP, esp. when you keep in mind additional luck+weapons-boni in that tree. There's a lot of stuff overpowered (basically all capitals) but I won't harp on that because there's logic behind it. Basically you've fastened up the game by generally decreasing technology & improvement-costs, lesser maint, and way more additional racial stats boni and powerful buildings. The net effect is that the game develops very speedy, and I think that's the right thing to do, Twilight was a lot slower than DA and if someone fires up the old game again it will feel fresh this way....
On the boni-improvement I mean for example the Temple of Righteousness or Hall of Empathy (lots more actually). You've added some bonus (pop growth, influence) but from the description it's not clear whether it is racial or planetary. Ok, this isn't different with most other buildings but the benefit of the old buildings is usually already known. I assume that most, if not all your newly introduced boni are racial, though I wasn't sure on some (eg +50% infl racial is heavy!)
Then there's the Dark Energy Lab, its description says it will boost "all our" military and production, is this civ-wide? Judging from the quote, I would say so yes, but is that even possible? AFAIK there's no racial production stat, but it should be true for the weapon-stat because there's no planetary weapon-stat either ... And there's further a +10% econ, which is totally omitted in the description.... confusing^^
Why not simply put thing clear there in numbers, either *planetary*- or *civ-wide* boni?
Maybe also a good idea for the boni you get when making the ethical decision. Some of these boni are mentioned quite clearly (+25% trade increase) some not (eg loyality bonus etc pp). To add, the neutral description on their Purchase Now reduction is faulty, because this will reduce not only ship buy but also their upgrade costs, and reduce the bcs of planetary imrovements.
re: Constructor spam
I've made some tests on this a few months ago and am very inconclusive about this in general. I don't think that it will work, I believe that the game will produce ships simply randomised and not by observing the state of its own game, ie. it doesn't ask itself the question if it will even need more constructors" etc... I do not know if this is true in the code, I simply assume from what I could observe in test-game.
I wanted to reduce the constructor-spam as well, and I was further irritated that the AI lost so many imprtant resource mines because he spents a boatload of constructors to increase mining before thinking on +att or +def on them. So I harshly reduced the number of modules so he could only put stuff on that was important (just like now) but the effect was that there were a lot of overflow constructors sitting at planets. It was funny because whenever a resource became free everyone did send all his constructors after it, that's how I became initially aware of the problem because you saw them pinpointing from all around the galaxy to it, even 20 turns after the resource has been reclaimed (the AI usually doesn't re-alter ship-destination before it reaches its goal)....
On the weapons-stuff:
This is a large and difficult (but very important) field, so I'd have to write a wall, but I'm pretty sure you know the topic very good by yourself.
You've doubled the the early wepaons (except the theory) and these weapons are important for the early game, and even more in less-big games (because usually there you don't need to research high tier weapons).
Increased weapons will faster drain the hitpoints from a ship, which means that defense absorption cannot work as good as it should. Defense and hitpoints do boost each other strength, while weapon fire does the opposite.
The problem of defenses in a fleet battle is that all-attack ships will cumulate their weapons but defenses cannot do this because only one ship will be fired at at a time.
Now this is alot of assumptions made on my end but I can give you some maths to exemplify what I mean.
From the game yesterday I take one medium hull, and with my current level of miniaturization build an all-attack ship and an all-defense ship using the same techs/tiers. Because the all-defense must have an attack as well I give the attack ship a small defense as well to make it fair. I get these stats: (using Singularity Driver & Duranthium)
All-attack: +36att +4def +20hp 370MP cost
All-defense: +8att +68def +20hp 480MP cost
Note: the all-defense ship is 25% more expensive, seems to have higher stats and should be better actually.
Let us now simulate a 1on1 ship fight using average rolls (* more to this later), ie simulating with 0 luck
All-attack +18att +2def
All-defense +4att +34def
Result: It very easy to see that the defense ship will totally absorb all damage and destroy the attack ship after 5 rounds of combat.
Now lets simulate a fleet battle, this time incorporating the build costs as well.
4 all-attack ships = (average) +72att +2def +80hp (1480 MP) vs
3 all-defense ships = (average) +12att +34def +60hp (1440 MP)
(some might say 4 vs 3 is unfair but the reality ingame will be exactly like this because you HAVE to take the cost into credit)
If you now calculate round per round you'll notice that the all-attack will totally drain the first target's armour, then its hitpoints destroying it, then draining their remaining fire in the second ships defenses.
The defense fleet will drain -10hp from their enemies. In the second round they're only 2 ships so they will drain only -6, giving the attackship still +4hp.
The attackfleet will again destroy one defender.
In the last 3rd round the last defender is killed while doing -2 damage, the attacker will barely survive with +2 hp.
Result: All defenders are destroyed, all attacker live.
The problem here is that the "invincible defense" is actually not really effective versus cumulative weapons-fire. I've done alot of calculating (actually since the time when I asked you guys about the fleet-module stuff a year or 2 ago) and the common consensus seems to be that defenses are NOT worthwhile at all when they cannot re-load during combat turns. In such a case hitpoint mods are always better.
(*) The ingame reality is even far more worse, because the game doesn't roll averagely but instead randomly, which is giving a huge bonus to the attackers but NOT the defenders. An attacker could roll high while the defender rolls low - in this case the defender is screwed. Butt the opposite scenario, defender rolls high defense & attacker rolls low attack nothing serious will happen, yes damage will be totally absorbed which IS actually already happening at ordinary average rolls.
If we take up the 1on1 fight example the defender needs at least 3 turns to be successful, although this chance will be astronomically low because the defender has to roll 3 times max attack and the attacker has to roll 3 times lowest defense for this to be accomplished.
In opposition the defender only has to roll a single time around -10% of his defenses while the attacker is somewhere above 75% in order to oneshot the defender. And this can happen during the length of the fight. Before I used a +50% Luck my defense ships always lost mysteriously vs obviously lower ships simply because of these unlucky rolls.
You can see from this example also why Luck is so important in the defense-design. No more oneshots, and thinking further, if we would further increase the overall hitpoints vs weaponsfire correlation, this would give defenses even more time to work, to reload.
Not saying now that this would take care about the fleet-battle defense demise - actually it would mitigate it but still the attack ships would win. But increasing weapons while leaving hitpoints as they are is not right track.
Still, this is only a single example and other examples using other weapons/def/hulls etc might bring another conclusion etc. Although I have to say that I consistently get likewise results. Usually ingame defenses aren't that bad... because if they don't suffice you research another tier and at some point they do their job. And of course, a ship equipped with tier 5 defenses will totally outclass a ship with only tier 2 weapons.
That, and the fact that defenses can easily be researched, is the reason why you can't simply buff defenses up. Then, in a 1on1 scenario total absorbtion will be reached quite easily and be a I-win-button, and if all damage could be absorbed hitpoints will loose all importance, ie. you can use Cargo Hulls instead of Medium Hulls. I did this successful in the unmodded TotA for some time, until longer games required ever bigger fleets and thus, defenses became too weak.
I've also played 3 months with simply all hitpoints increased by 100%. Combat rounds are increased, firsthit advantage is reduced (sad for Arceans), even tiny hulls can benefit from a single +10% hitpoints boni, repair becomes important (this used to be an important stat in GalCiv1) but the general problem here was that the additional hitpoints received from level up were too excessive.
As of now, I believe increasing hitpoints to a 1.5 ratio as normal (cumulative through changes in the hull base points and some additional hipoints racial stats) while keeping highest tiered weapon and defenses out of the game is the right track, but I need more testing time in my personal mod.