It's still worth the faction happiness loss, to be completely honest.

Those are rather inefficient ringworld segments. Slot your refineries on generator and agricultural segments, and then fill out dedicated research segments with advanced labs. The overall tech output will be much more efficient, and if you tally up the overall sprawl cost it will be much lower for the same research output. I've done a lot of number crunching on sprawl penalty, and for late-game empires it turns out that minimizing sprawl is just as important as total research output (if you can research enough of the +15 admin cap techs to eliminate your sprawl penalty entirely, research output goes crazy)

The vibe I am getting from the discussion is that many of you would consider habitats if they were merely a rare tech instead of an ascension perk. I could see myself using them in this case as a stop-gap measure sometimes or for the option top cram fortresses onto them, but not very frequently. During the early game I need as many pops as possible, so that might be another use for habitats, but then again, during the early game I am also starved for minerals, so that I much prefer settling a planet instead.

To me it boils down to this: In the early game I do not have the ressources for habitats, once I get a decent economy going I am usually able to unlock the arcology perk, after which I go for galactic wonders. The only time where I might be looking at habitats is between the point where I get an economy going and the point I am able to construct ring worlds. The thing is that during this phase I have other, better options like arcologies and I am certainly not going to invest an ascension perk into habitats if instead I could get arcologies.

The empire sprawl thing is an interesting mechanic, but I am not sure how relevant it is in the context of habitats. It used to be that you would get a penalty for each settled system, but now you get +2 sprawl for every colony and +1 for each district. I would need to check again, but as far as I recall habitats do not give me an advantage when it comes to output versus empire sprawl. Please correct me if I am wrong. Even if they were a way to effectively reduce empire sprawl versus output, during most of my games I research Administrative Capacity to levels beyond 100, with research time per level being around 12 months with egalitarian empires. Other empires I have played around with, but I have difficulties finding a viable late game research strategy without the egalitarian option of utopian living standards. So while egalitarians may cram fortresses onto habitats for fleet capacity and research, I can not see any uses for habitats in other empires.

In fact, I once conquered an AI with an absurd amount of habitats, but they were such a drain on my economy that I just went on to resettle everyone to my ring worlds and arcologies and proceeded to wipe out the remaining habitats with my colossus. Fun fact: I was not aware of the diplomatic penalty this course of action would incure from other empires. They mad, bro.


edit/
in fairness, I can not tell you why I did not go the route of cramming habitats with fortresses instead. Could be it never crossed my mind, but I guess part of the reason I resettled pops was because I needed to reduce empire sprawl and I had many better districts on my arcology with lots of housing and jobs in amounts that habitats could simply not generate efficiently. So... because districts cause empire sprawl, and because you will get much more out of districts on arcologies and ring worlds, this is another way that habitats are bad.

By "efficient" I meant less population required for the same research output. They also don't cost any special resources and the research districts cost 2 energy credits per 3 researcher jobs vs 4 energy credits per 2 researcher jobs from research labs.

You are right about the sprawl being better on Ringworlds, especially if you upgrade the laboratories (You can see that in both images where the Ringworld goes from 6 districts to 2 while providing the same research output). I'm sure this is amplified further since I think roughly 5 or 6 habitats are needed to generate the same output as a planet or as a single Ringworld segment focused on pure research, making the sprawl quite significant with the addition of +2 colony sprawl penalty for each habitat.

EDIT:Just read your post OP; I think even if habitats were easier to obtain in the early game, I don't think I would like to see them become disposable late game as a result. If possible I would like to see habitats at least excel at something that makes it worthy of being a "Mega-structure"...although completely my own opinion.

The issue is how much sprawl it costs to support researcher jobs. While research labs don't cost sprawl on their own, you do need city districts to house the researchers, and elsewhere you will need city districts to house the workers who make the consumer goods and exotic gasses that are used by the researchers, and then there's some overhead for the planet itself. When you add it all up it comes out to about 4.5 researchers for every 1 point of sprawl. When you calculate it out for research districts in habitats, it's only about 1.5 researchers for every 1 point of sprawl.

I've done a lot of spreadsheeting, and have basically concluded that for very large empires sprawl management is just as important as how much science you're outputting. In particularly extreme cases, you can actually end up reducing your science output by adding more researchers if you added too much sprawl in the process!

I've been doing much the same, although in my experience you can usually just go to war and claim enemy worlds to fill your expanding admin cap, so it's not like you've got room to waste. And if you're doing this you do not want to go over your admin cap even a little (you're literally better off abandoning territory than going over admin cap at these scales). As far as science on habitats go, unemployed pops are way, way better than science districts.

As for producing science without utopian abundance, the trick is just to build lots and lots of labs.

There shouldn't be a penalty for doing this, not unless you purged them. Just resettling them off the habitats (which you should definitely do; the AI builds way too many habitats) doesn't piss anyone off.

They weren't mad about the resettling part, but about me neutron sweeping their habitats. I did not want to resettle everyone and then create and apply new species templates for all the pops, also all my worlds were filled up already and the ring world would take another few years, so... I basically did what the Mongols did to China. China is still very mad about that. And so were the other empires. And then the Scourge arrived and even more people died. All things considered, the galaxy was still in a better place than Warhammer 40k, so I consider it a success.

A major points in favor of habitats I read elsewhere: If boxed in early territorially without access to reasonable colonizable planets they can give the tech leg-up to be turn the game around even at the cost of an AP.

To be honest, I misunderstood their use and this is also a feature that would make voidborn a relevant perk: if attached to an un-colonizable planet, they function as they are now, but you can attach them to a colony without incurring extra sprawl to add 4-6 additional districts to it. That would also close the ridiculous gap between regular colonies and ringworlds... And I like the RPG component of habitats for a space-game!

I guess if this thread is bumped, I'll just drop this here

So habitats will be getting overhauled in the next major update.