I agree with the general thrust of the argument that food is relatively unimportant, but this seems to be super ignoring the opportunity cost of the hydroponic farm. A better comparison is 5 organic pop for 7.5 energy because if it wasn't a farm it could be a power plant.
That opportunity is lost if the power plant built in the stead of farm to support a synthetic population; the farm supported more pops for less energy cost whilst providing the additional growth benefit, coupled with the efficiency lost in synthetics being unable to use any food bonus tiles. 2 food tiles are ubiquitous; if you must use it for energy or for food, you would as an organic rather use it for food. The power plant lvl IV with maximum building efficiency (so energy nexus & stock exchange) would yield 5 x 1.3 energy, or 6.5 energy. In a purely synthetic state this would mean two such tiles would provide for 13 synthetics. An organic state however would be able to use that food, and thus with hydroponics lvl IV and an agrarian pop with nutrient replication would yield (5 + 2) x 1.3, or 9.1, meaning two such tiles would provide for 18.2 organics, at the cost of 5 energy - which is just under the difference between what agrarian farms & power plants can support in their respective populations.
But if you go for the pathways available to organics with access to slavery or harvest, things get a little spicier. In the previous example, food production is more efficient, but the resources gained in the competitive advantage are offset by the energy cost, if you for the sake of heuristics ignore the effect on population growth gained by the expenditure of potential energy.
An agrarian slave working on a lvl IV hydroponics farm with the slave processing facility & nutrient replicators on a 2 food tile makes 7 + 50% or 10.5 food per tile. This would be equal to a synthetic Empire utilizing maximum efficiency power plants on a 2 energy tile, but for the organic state - the organic state is capable of utilizing the maximum efficiency of both food & energy tiles, an option unavailable to purely synthetic ones. This is why the opportunity cost is not comparable, because the opportunity is unavailable to the purely synthetic state.
While planetary modifiers affect this about equally (with hazardous weather being just slightly inferior in causing -5% happiness alongside +20% energy gain, whilst lush causes only positive effects with +10% habitability & +20% food), there is the one notable modifier of titanic lifeforms being harvested available - which provides +100% food to the planet. This modifier is broken, there is no reason not to take it (unless it's really early game and you can't defend your planet against titanic lifeforms. Or perhaps you want them working for you). There is no comparable modifier for energy.
Not to be understated is the paradise dome's +5% happiness. +2.5% bonus to every resource production will easily overcome any gain from having built a power plant instead.
Processing as livestock is notable for being hilarious, but it's more of a novelty than a useful strategy.
Tl;dr; it always comes back to the problem that organic states can use all the advantages of synthetics alongside all the advantages of organics, whilst synthetics cannot. The addition of machine worlds to the roster of synthetics has given them more purpose to exist, but imo synthetics still need more exclusive bonuses to justify becoming them, instead of just using them or starting as them.
All that and as you yourself say, late game you can often end up with a LOT of energy, which makes this comparison tip even further towards energy, since this acts as an efficient dump for it. Gardens are good for sure. Although, as you also say, the unity eventually ends up useless! At which point you have to question if the happiness is worth it. I think the answer is generally going to be on some planets and set ups yes and some no.
The happiness is always worth it, there is never an instance where you do not want every single one of your pops to be at 100% happiness, because they'll be drifting towards your ethics and giving +20% to every single resource production. The answer is to every planet: Yes, absolutely it is worth it. As for energy dumping, it's much better to dump that into more fleets than pops
I hope they eventually add more late game uses for unity. I recently played a mod (forgot the name, but it's fairly popular afaik so probably you guys know it already) that added late game buildings that gave a ton of science but had huge unity upkeep costs. To the point where it wasn't possible to put one on each planet. It stuck me as a really good idea and use for late game unity. Just, big things in the late game that constantly drain it.
Also the addition of capital ships which cost unity to maintain in the more ships mod. Having unity be like that is just a great idea
The current planet crackers might be an example of an area where this could be applied. Maybe instead of a hard limit of 1, they cost somewhere between 300-500 unity per month to upkeep? After all, keeping up the literal ability to destroy a planet seems like it should be somewhat decisive and worthy of constant evaluation.
Would probably also be cool then if you could get into negative unity, instead of just 0 unity, indicating the appearance of disunity within your Empire
This kind of reminds me of the various dominions balance arguments. My guess is that different sides of this argument are playing with different galaxy size and civ density options.
Synths give you a big power spike when you convert all your pops, but they make it somewhat difficult to expand your pops. They also hurt you in the habitat game simply because the minerals you spend on robots could also be spent on habitats.
I'd say Synths are pretty good for populating habitats though once your industrial base can accommodate the building costs. Delicious research and energy production synergies with synths pretty good
The perspective I'm coming from is someone who plays single player and notices that most games get "resolved" early on when a couple obvious super powers emerge and start snowballing. So as the perk path with the more powerful 1st level perk, and the more "front loaded" 2nd level perk, I would argue that synths are better, since they more rapidly get you to that point where no one can match you and you snowball easily. As long as you have a good chunk of the galaxy already fully populated by the time you ascend, the awkwardness of building robots isn't going to put you back as much as the huge production boosts and the conversion of your existing leaders into immortals. Maybe MP works differently than SP and games go long, I dunno.
The first psionic perk makes all of your pops give +5% research, access to the psionic leaders & the shroud, the first biological perk gives you +2 trait points and cuts modification costs, with both costing you cheap society points instead of valuable engineering ones. Both of these offer better than +20% habitability from cybernetic and offer better long term research options, with the added benefit of not wiping out all your species specializations, and with psionic especial mention must be given that there no need to establish hegemony early on. Thus for example if you reach the point where building robot pops is not a hassle, you're already at the point where no one can stop you, and the ascension perk can be better spent on the late-game focused perks. If you have not, then psionic & biological offer the mechanics needed to maximize research; especially with whisperers in the void, or the option to call the apocalypse, double your everything & make bid for hegemony/watch the end kill your enemies.
Of course, an additional component of my perspective is that I don't play the game one-planet because I think that's an exploit and I might as well learn to play without it because its probably going to get patched. So I tend to expand out over the course of the entire game. So to someone who plays one planet, pop growth speed is very important, they have few existing pops that they need to modify and a lot of research points with which to do it. And of course, they get to that magic 4th ascension perk way way way faster than a more conventional strat would. But playing more normally, by the time you'll have a 4th ascension perk you have a lot of pops, so that power spike you get from converting all your weak organic flesh into ultraproductive synths is more pronounced.
Playing normally, one should not run into any trouble getting to the 4th ascension perk by prioritizing all the unity increasing traditions in the right order, relative to your expansion. One planet is wonky as hell, and actually decreases in efficiency compared to a core-sector strategy when you start getting T3 & T4 research, and a one system strategy is mostly just viable for Sol because it has Mars and lots of Habitat space to build additional research labs, but I think that's another issue entirely lol