I remember tinkering with nearly all the planets I gained in the original Master of Orion. That's something I don't do to nearly the same degree in modern 4X games.
There's a reason I keep an installation of Master of Orion 1 on every computer I use, and that's because despite being very early to the genre it was as close to perfection it ever got. By making planetary development a couple of sliders but with a depth to them that makes it worth paying attention to, they avoided the usual boring optimized build order routine of 4X games that take inspiration from the Civ series, like Master of Orion 2. Planetary size had impact on your max population, and different climates made planets inhabitable until you could figure out the technology, and later on you could terraform them.
Increasing complexity is what led to Master of Orion 3. Audiences thought they wanted MOO2 but with a larger map that took up more than one static screen, something that made it feel like you were ruling a galactic empire with all of what that entailed, hundreds of space colonies, but when you scale that up and put it through the Civ formula you either get a game that plays itself or that gives you carpal tunnel syndrome, which wasn't helped by how many submenus MMO3 had. Take that even further and you might end up something like Distant Worlds that is more of a simulation than it is a proper game, which can be fun for the toy-like boy's first steam engine or antfarm aspect of it, but unsatisfying if you want something more akin to a boardgame.
It's not just about complexity, simulationism and automation, but meaningful decisions. In Suzerain, which isn't a 4X game or even a real strategy game, but more like a visual novel for people with an interest in geopolitics, a big infrastructure project is a big decision, what it will be, what entity will be responsible for the building of it, and in terms of power structures and alliances. Do you scratch the back of oligarchs that might fuck it up by being cheap? Do you hire a new company that gets things done the fastest but is unproven? Do you allow the slow and corrupt state construction company deal with it? This is an interesting decision with all sorts of interesting implications with long and short term consequences.
In MOO the few sliders you have might not seem like much at first, but they will decide if a planet is going to be a defensible place in the future, if it will be an industrial zone dedicated to producing your interstellar armada, or perhaps a planet of the best and brightest gather for military R&D. Every adjustment you make to the sliders have a huge impact on your empire. Then there is the ship designs, where a newly researched weapon could change the course of a war. Suddenly you might have death spores and can purge alien worlds of their populations with ease and no longer have the need for large invasion forces on the ground. Because this was such an early game in the genre what was implemented was justified by the impact it had on the game.
Fast forward to Paradox's entry into the space 4X genre with Stellaris and the gameplay now consists of micro-managing hundreds of worlds with buildings, while getting worthless popups each turn with "story" that gives you 1% bonuses and tiny material boosts to your silly Warcraft 1 styled resource bar or space mana. It's one of the worst games of the genre I've played, much worse than MOO3 even, and that's because nothing you do in that game ever feels impactful at all. You might play as a race of undead parasites attempting to take over the living of the galaxy, or a machine civilization, or the Star Trek space federation and it all is very samey. Meanwhile the race you played in MOO had huge consequences to how you played the game. Playing as quickly multiplying space frog-dinosaurs was radically different from playing as grey techies or the infiltrating and spying cloaked assholes that'd be leeching off you when you played as the greys or blowing up your factories if you were the robots expert at mass production. Because the systems were so simple any change to that was massive.
At the end of the day video games are supposed to be interactive experiences, and the more a game makes you think about your decisions, be it an RPG character build, a story choice, or how to manage a space colony, the better it is. Typically the more you scale things down the more impact your decision will have. If you have hundreds of points to spend to give you minimal % buffs in stats or attacks this becomes a non-decision and doesn't respect the player, but if you have one or two points to spend, and this will add +1 or +2 to a D6 or 2D6 diceroll under task resolution checks of some kind, this is huge. The same is true for any genre, 4X, or space exploration games. Landing and walking around on each planet in Outer Wilds is radically different, landing on any of the 100 planets in Starfield feels exactly the same, so why have the 100 planets, what's the point? Again, it's about respecting the player, making interactivity meaningful.
Believe it or not, Elite's galaxy was actually procedurally generated, but it used the same seed for every game.
It was a technical requirement back then for that type of game, Starflight had a procedural galaxy to explore, that was static and didn't very between playthroughs, but had to be generated on the fly. You could land on planets, interact with the wildlife and things like that. It got its character from the aliens that populated the stars, with very varied attitudes and distinct personalities, and gave the world distinctness more by the lore and story you'd find on distant planets moreso than through the depictions of the alien worlds themselves. This was in 1986 though, and the focus was on the scale and fantasy of having a crew aboard a spacship, landing on alien planets and exploring them, and meeting strange new life. The planets in Elite are more like the towns and colonies in Sid Meier's Pirates, ports to trade with, pit stops on the road, since the focus was on space itself and dogfights, and I don't think you can compare that to what Starfield is attempting.