I think I wanted to like Stellaris but... I just can't. Its a very wide game, but the "puzzles" and "learning" seem very shallow to me.
Before talking about the bad, I probably should talk about the good.
* Fantastic diplomacy / ethics system, providing plenty of roleplay available for a variety of sci-fi concepts. Xenophobe (kill all aliens) vs Xenophile (aliens should live together in peace!). Spiritualist (psychic priests anti-robot) vs Materialist (pro-robot cyborg races). Pacifist vs Militarist. Lots of random generators to enemies and a relatively simple to understand ethics system to quickly understand how your neighbors think of you, and what they want out of the galaxy.
* Good ship design mini-game. There's lots of features to choose from: Corvette, Destroyer, Cruiser, Battleship. Each with 1x to 3x sections, each section providing different slots, each slot providing different weapons. Its ultimately a large-scale rock-paper-scissors match between long-range kiting, missiles/strike craft vs point-defense and flak-cannons, energy shields vs armor... and guns vs lasers. It takes a long time to get used to all the options (and which ones to rush for in the tech tree), and it feels good to make decisions (mixed weapons provide more variety in combat. But focusing on all-energy weapons will help focus your research tree, even if it opens up a weakness vs shield based defenses). Especially since the 3.6+ update (December 2022 and later), Stellaris's ship design is deep and thought provoking as a design question / optimization problem. And ties well to the rest of the game.
* Sci-fi theme, with plots shamelessly stolen from other sci-fi culture. This is a good thing! United Nations of Earth is basically Star Trek. Commonwealth of Man is basically the 40k Imperium of Man. Etc. etc. With such a broad-reach of concepts, you can roleplay from any Sci-fi genre from Starcraft, to Star Trek, to Star Wars, etc. etc. Or participate in the actual lore of Stellaris itself.
---------
Everything else? Ehhhh... not so good. I found the planet-design minigame to be shallow and too easy (despite having to memorize huge amounts of data to come up with a strategy... it comes down to aim for +1.5 pop bonus. Try to grow faster with +pop growth, and specialize your planets). Its a lot of micromanagement that's... not hard to do... but requires a lot of clicking. Ex: unemployed robots cannot automatically resettle before the "Synths" tech. So you gotta go from planet-to-planet and transfer robots every couple of months to solve your unemployment problems. These.. should have been handled by the computer automatically, somehow. Its dumb, pointless work that distracts from the game.
Clerks, a particular job type, are awful unless you're explicitly going for a trade strategy. Every time you make a planet, you need to go into the jobs section and disable all Clerk jobs. (Its better for workers to be unemployed, then resettle elsewhere in your empire, than for those workers to work on a near-worthless Clerk job). Etc. etc. Lots of clicking for what amounts to an extremely easy decision. (IE: Are you a corporate overlord? Okay, clerks are fine, don't disable them. Not a corporate overlord? Disable all clerks throughout your entire empire).
But the game is just _filled_ with these low-effort, low-thought loops of busywork. Oh, a leader died (especially now with leader caps in 3.8, this is huge). I guess I need to go through the list of available leaders to hire and think about which bonuses are best for my sectors again. And now that nothing matches well, I'll have to revisit and re-read the list of next leaders. Oh no, a leader developed a negative trait. I guess I gotta dismiss them, and hire a new leader without a negative trait. Etc. etc.
And its not even that hard. You get +Research leaders to lead your +Research planet. Go get +Farming leaders to lead your +Farming planet. But its a lot of busywork reviewing the lists of leaders looking for a +Research bonus governor, because these attributes are random. And there's penalties (in 3.8) for having too many leaders, so you gotta fire leaders to make room for these other leaders. You gotta hire temporary suboptimal leaders while waiting for the leader-pool to refresh. Etc. etc.
--------
"Events" are cool from a lore perspective. But the decision-making is absolutely insane, especially given the different levels of power available. For example, the Ameoba Flagella event provides two research options that sound the same. But the first option leads to +5% evasion, permanently, for the rest of the game applied to all ships you'll ever create. The other option leads to +33% damage vs space ameoba, which is... kinda worthless. These things are the easiest things to kill in the game, so there's no reason to ever pick this option.
But you don't _know_ what these options path lead to at the decision point. The only way to know that the +5% evasion choice is better is if you reverse-engineer the game and read the scripts in the Stellaris data directory (or maybe consult the online community, who has figured out the 'optimal choice' on a lot of these events).
----------
There is a good game trying to come out here. But... all the pointless busywork, the unpredictable event (and event bonuses, without much hints in the text to which choices lead to better/worse options for your builds), makes the game a chore to play.
If the event-system were more focused on lore (which could be as easy as removing some of the exceptionally good bonuses, like +5% evasion permanently for the rest of the game), then that's cool. Heck, even minor bonuses (a one-time +500 energy bonus associated with some star-discovery event or something) is fine, and these constitute the majority of events in this game. But knowing that there's permanent event bonuses that can only be taken if you memorize events and optimal choices to lead up to those events? That's... not fun. That's just me, being a dictionary / reading the Wiki looking for the optimal choice now. You can't hide god-tier bonuses inside of the event system like this.