No improvement is truly good without the appropriate civics. If you use State Property, both become quite good... in fact, with Caste System and State Property you should not work farms and mines:
1 farm, 2 mines gives you 2
6
1 workshop, 2 windmills gives you 2
6
2-6
They may be worth using without State Property. Corporation food can easily render additional farms excessive, watermills can certainly compete with non-optimised cottages. Depending on the local food surplus and whether you run Caste System, even workshops may be worth considering.
*
Workshops and watermills aren't spectacular in the most *powerful* lategame economies that break corporations and/or rushbuying... but those require a lenthy set-up and/or investing in butter when guns are more tempting.
Beelining Communism and using them extensively is a practical alternative... appropriate, the underlying thought is 'that'll do until I win'.
The future payback of workshops is smaller, but the payback within turns which are relevant to the game tends to give them a lead, if you can get there in time.
Generally non-riverside non-financial cottages are a tough sell throughout the game; they take quite some time to compete with raw spec yield or alternative tiles. Some capitols definitely merit them, as do very powerful mass riverside cities that don't quite have enough food to be viable for globe. No matter what, GPP is necessary for those who aren't just lazing through the game (so often if I get enough early land I'll just cottage it and sit on it until it's time for tanks/bombers or something).
Anyway OP, with SP and the right techs workshops are among the best improvements in the game and the only set up costs is the worker turns to build the workshops and having the requisite techs/civics. Watermills in particular are incredible with SP and a levee.
One improvement I think people under-use is the lumber mill. There are often non-riverside forests (plains or even grassland) that really aren't viable tiles in the early goings for cottaging, and
cap might not merit a non-riverside grassland farm (and certainly working plains at that point isn't hot). On the flip side, each pair of forests is worth
, even with the lumbermill on later. Sometimes that health is hard to quantify, but sometimes it isn't! If you can't pull the trades and would otherwise be unhealthy if not for the forests, you can think of them as aqueducts or as "farms" in the sense of providing food through health. Later in the game these things tend to help productions cities immensely; they have the same raw yield as mines but with a more favorable health situation in cities that will struggle with it.
I'd still like to see them slightly earlier so that the chop vs mill dynamic becomes very real, but w/e.
IMO too many of the more experienced players on the forum lately hate on cottages a little too much. Riverside cottages and FIN cottages in general are pretty solid, as long as the player comes up with some decent other city specializations (GPP, heroic epic for example). They also work well for players who are deliberately allowing the game to go on long enough for their ROI to make sense. Cottages placed in the first 50-100 turns are also mature in time to be competitive with alternative improvements. Workshops only dominate cottages mid-game on. At the very start of the game on say a MC beeline, a workshop is FAR WORSE than a cottage, having a NEGATIVE payback (-1f +1h) while the cottage adds a little. Printing press villages do pretty darn well vs caste/guilds workshops...really anything but SP workshops. How many players reach communism before their cottages reach that point in their initial settling phase?
The above doesn't even take into account bureaucracy bonuses for hammers/commerce or that free speech to boost towns comes alongside one of the techs players gun most consistently. Early cottages work well if you don't overdo them, although going democracy ASAP to emancipate suggests a long-term investment beyond initially funding expansion + some tech.
Most of this isn't directed at Iranon, but just in general. The growth time on cottages is a direct attempt to balance the fact that they are viable options MUCH sooner in the game than workshops/watermills/windmills. GPP is good but you can only get so many and usually after 1 good NE city you won't have much ROI on more GPP investment in other cities. Non-rep specs are...not so hot compared to cottages. Mines do a lot better, if they happen to be available.