is it useful to be really aggressive with pastures?
Not particularly. Pastures give diminishing returns. I've never had more than 100. Definitely build them, but I find aquaducts to be more important to push.
how much of particular buildings to get
In general, you don't stop building buildings. How many buildings you can build is limited by your resource caps. When your caps increase, build more. If you, say...go out of your way to build lumber mills, and then once you have ten you're incapable of building any more because your resource limits won'tallow it, great! You're in a good place. But if you then stop building lumber mills because of that, that's a mistake. Once your resource limits increase, build more lumber mills. There's isn't usually a target stopping point.
There are, however, some buildings with heavy diminishing returns. Like pastures. And some that you only need for afk play. Like steamworks and mints. Some that simply so expensive that they're not worth continuing to build past a certain minimum threshold, like temples. And some that eventually reach a point where it stops being very important to build more because you don't particularly need more of them, like the catnip buildings.
Workshops. I have yet to find any point of the game where building more workshops wasn't worthwhile. However boring or pointless it might seem to wait an extra 20 minutes for a mere 6% when your workshop multiplier is already in the several hundreds, workshops pay for themselves very quickly. Especially later in the game when you start using crafts to make crafts to make crafts. Crafted (not traded) blueprints benefit from the workshop multiplier four times. Fur becomes parchment becomes manuscript becomes compendium becomes blueprint. 1.06 ^4 = a 26% increase. The compendium griding in the early middle part of the game is also especially bad. Workshops help make it less painful. And through most of the game, scaffolds are used for a whole lot of things. Build workshops. Workshops are good.
Lumber mills/mines/aquaducts: Not much to say here. These increase your basic resources. More is good. Yes, aquaducts become less important later on as catnip becomes abundant, but they're also the primary thing keeping your kittens alive through cold winters. If you have too many aquaducts, it just means you get to turn more catnip into wood. If you have too few aquaducts, all your kittens die. Mines are cheap, and are easy to keep near capped. Go ahead and do that. Lumber mills are more tricky because the iron requirements are so high, but wood remains important through nearly all phases of the game and it's generally something you need a whole lot of. Continue building them throughout the game. Balance lumber mills with your iron needs. Be aware that the zebra trade gives especially good windows of opportunity to build lumber mills without having to wait hours for iron, because they tend to throw a lot of iron at you. Take advantage of that.
Smelters: Early on coal is a huge bottleneck. Smelters with the right upgrades produce coal. And then much later steel is a huge bottleneck. But there's a big delay between getting steel and it being a huge bottleneck. There may come a time where you're looking at your pile of 500 steel wondering what to even do with it, and it might seem unimportant to even bother continuing to make steel. And then much later you'll realize that oh, you need 50,000 steel to progress. Granted, that 50,000 steel bottleneck won't happen in your first or second run. It's not until you start exploring space. But it does happen. Having a lot of smelters running with workshops to give multipliers helps you collect that steel as you go so you already have it when you need it later. Also, gold can be a huge bottleneck for a sizable portion of the game, because it tends to be the limiting factor in trades. There are some points of the game where trading is the only way to get certain resources. If you're ever sitting at mineral cap and don't know what else to do with them, build more smelters even if you don't think you need them. Later on, you might. And then even later after you've kind of forgotten about them, you'll get an upgrade that allow them to produce titanium. You'll never regret having more smelters.
Observatories: On a typical mid-game run, you'll want on the order of roughly 750 + 75 + 50 + (25 * 243 / workshopmultiplier) starcharts. Two thousand or more. Starcharts come from astronomical events. Observatories make them happen more often, and they give a chance to auto-collect when the events happen. If you're playing actively, observatories give you more opportunities to click. If you're afking, observatories increase the collect rate of events when they happen. Even if you don't need them for research, it pays to build lots of observatories for starcharts, typically up to the point that you get the 100% auto-collect upgrade and end up with thousands of them sitting around. At that point you can stop building observatories. But even then, don't dump all your starcharts into trade ships, because space mission require hundreds to low-thousands each.
Trade ships: The target for trade ships is 243. It's not essential that you get to that point, but that's the "sweet spot" for zebra trade efficiency. You'll hit hard diminishing returns beyond that point, and it may be better to simply stockpile starcharts for the later phases of the game.
Paragon: Paragon hits diminishing retuns at 150. Having more is not bad. Having less is bad. You do want 150 paragon. But if you have more, spending it is probably a good option.
kittens: Kittens give you the ability to generate bsic resources, but kittnes beyond 5 reduce happiness. Happines is applied as a production modifier. There was some discussion of the math in
this post, but short version: there may come points where having more kittens
reduces your overall production because they're sad and mopey from how overpopulated it is. That point is highly dependant on your stats and it's difficult to give targets. In general, don't let your happiness modifier get very negative. This becomes less of an issue in subsequent runs. Build amphtheatres to help this.
Ampitheatres: these overate on mulitiplicatively diminishing returns. Every ampihthreatre you build reduces the effect of the next one. The penalty can never reach zero. 1 ampitheatre changes the multiplier from 100% to 95.2%. 10 reduces it to 61.1%. 20 changes it to 37.3%. 50 changes it to 8.5%. 100 changes it .7%.At some point you can probably stop building these.
Pastures: Diminishing returns. The penalty can never reach 0. Yes, build pastures. Don't neglect them. But at some point, more will stop making a difference. Especially when you get to the point that your kittens pretty much can't starve no matter what you do.
Mints. Mints are useful for collecting fur overnight. If you afk overnight a lot, mints will be more useful. If not, mints are useless. Possibly even detrimental because they consume gold, which you'll often have better uses for. They also consume catpower, and most of the time it's far more efficient to use catpower manually. If you do afk a lot, mints become more useful. Personally I've never built more than 5, and I haven't felt like I'm missing out.
Steamworks. Like mints, these are mostly for afk play. If you;re mostly an active player, steamworks are not very important. A couple for energy generation during a relatively small window of the game is all you'll need them for. 2-4 is probably plenty. And don't rush to build them. If you afk a lot, steamworks become more important, but even so, they require upgrades to be useful. In particular the "make manuscripts" and "make plates" upgrades are great. But steamworks require blueprints, which tend to be in short supply, and I've found it difficult to build enough steamworks to be very useful without gimping myself. I wouldn't go out of my way to build very many of these. I usually stop at 6, and don't even bother building your first one until you either have the manuscript upgrade or enough steel and blueprints that the cost is insignificant. Also remember that steamworks consume coal while they're on, and unless you're afking, you probably have better uses for it. Building and powering on a steamworks the moment you unlock them will probably hurt you because of the coal usage.
Log houses/huts/mansions: These are a mixed bag. Yes, you need them because they give you kittens. Kittens give you production. But the more kittens you have, the unhappier they become. Unhappy kittens are unproductive kittens, and there comes a point where having more kittens reduces your production rate because of the happiness modifier. To find out if adding more kittens is beneficial, compare:
current kittens * (happiness / 100)
(current kittens +1) * ((happiness -2) / 100)
If the second is bigger, add more kittens. If the first is bigger, don't add more kittens until you're ready to reset. Then get all the kittens you can because production won't matter after you reset, but the karma and paragon bonuses from having more kittens will.
Unicorn pastures. Cost increases on these rapidly hit diminishing returns. Somewhere in the 10-15 range it may be better to stop building pastures and instead sacrifice the unicorns for tears. More than that and you can very easily get to the point where buying more will never be cost effective over the entire lifetime of your run. Note that unicorns are late-middle game commodity, and in your first couple runs you simply won't be able to do anything useful with them other than just have them.
Ziggurats: Like unicorn pastures, you'll simply be unable to do anything useful with these your first couple runs. Unlike unicorn pastures, they require resources you have much better uses for. Go ahead and build one to play with and see what it does. Don't bother building very many until you're in space, spending paragon, and the idea of casually spending a couple hundred blueprints doesn't bother you. Don't build a bunch in the early game because you have no other use for the blueprints, only to discover that oh, they do nothing for you and you now need 100 blueprints. Oops.
Magnetos: These exist primarily to keep your biolabs and oil wells powered. But if ned more oil wells, it's probably more because you want to increase your cap rather than to increase your production rate. Once you start exploring space you'll want a lot of oil, but not terribly often, and if you need more oil you can adjust your power generating/consuming buildings temporarily. Most of the time, you simply don't need very many oil wells running. It's not necessary to have enough magnetos to keep everything running at all times. Note that you can't turn biolabs off, so you will want at least a couple magnetos. But don't go out of your way to build them. They're expensive and require resources you might have better uses for. Yes, they give a production bonus. It's nice. Don't go chasing that production bonus only to realize that you spent all your blueprints and need to spend the next 2 hours grinding fur and trading to get them back.
Trade posts. Yes, trade posts are good. You do want some trade posts. But, they become very expensive in terms of gold, and you need that gold for doing trades. At some point it stops being cost effective to buy more tradeposts to give a tiny bonus when you could instead use that gold to do 20-30 actual trades. Trade posts are not workshops. Don't hurt yourself buying. Also, be aware that they don't benefit what will probably be your most important trade: titanium from the zebras. To improve that, build trade ships. Much later, trade posts will help with the uranium trae from dragons, but...not by very much. Yes, build trade posts, they're nice to have...but don't build so many that they take up all your gold. You need it for trading.
Calciners: Yes, calciners are helpful. But they also cost a lot of blueprints and consume power. And if you have a decently sized trade ship fleet, it's a lot easier to get titanium from the zebras than from calciners. They're mostly an afk overnight kind of thing. Yes, build a few. But don't build so many that you use up all your blueprints to decrease your titanium production-to-cap time from 6 days to 2 days, only to then discover that you need 100 blueprints and don't have any. Calciners can be more or less helpful depending on your stats, but don't always assume that more is better. Very often building more can hurt you more than it helps.
Temples. These are super expensive and you have better uses for the materials. But you'll need to build them to increase your faith cap to useful levels. Don't go out of your way to build temples. Build them when you need to increase your cap and when you won't miss the gold and manuscripts.
Biolabs. Not only are these expensive, they consume energy and you can't turn them off. This is a nuisance. But at some point you'll need them to increase your research cap. Build them only as needed.
Oil wells: When you first unlock these, they don't really do anything. They produce oil, but you can't use oil for anything. And they consume energy. Go ahead and build a couple, but you unlock them much earlier than they benefit you. Don't get caught in the "build oil wells to be able to afford magnetos to be able to power oil wells" circle. It consumes reosurces and doesn't particularly benefit you. You will need oil wells in stupidly large quantities once you start exploring space. But up to that point, you only need as many as you need to keep your calciners and magnetos running. Not much point building more than that.