One of the core mechanics in Civ 6 research is Eurekas. These are specific events that immediately grant you half of the science or culture required for an advance. For example, if you build a Water Mill you get a Eureka for Construction. When you kill a unit with a Knight, you get the Eureak for Military Science. Eurekas can require kills, owning units, buildings, wonders, diplomatic actions, and more. In the early going this is really cool because many of the Eurekas make a lot of sense, and it feels exciting to see advances in specific technologies or policies because you are already doing that thing.
However, this mechanic leads to some strange situations. For example, I don't usually build Privateers. They are middle game ships that are invisible unless you are beside them but I usually don't find time to make them. However, if you own 3 Privateers you get the Eureka for Electricity, which gains you 625 science. Each Privateer costs 280 production, so you invest 820 production to get 625 science. That isn't a particularly great rate of return... but you still get 3 Privateers that can run around exploring or shooting enemy units! The first and second Privateers aren't especially good unless you really need boats, but that third Privateer is *ludicrous*.
This mechanic is found all over. Workshops are pretty horrible buildings. They cost 195 production to make and only make 2 production a turn. 97 turns to return their cost is wretched. But you get a Eureka for owing three Workshops, and that Eureka is worth 422 science, so while the first two Workshops are awful that third one is the best.
This mechanic is a massive driving force in maximizing your power. There are times when you can't get a Eureka because the AI won't cooperate or you are under attack, but the great majority of the time when you look at the cost of whatever it is you are doing to get the Eureka you absolutely must make it happen. If a unit or building is even close to reasonable in a vacuum, then when it gives you half of a technology in addition it is better than anything else you could be doing.
This leads to strange consequences. Spearmen are basically junk units, good for nothing. But there is a Eureka for killing a unit with a Spearman so I build exactly one and figure out some way for it to get a killing blow.
This gives the game a jarring feel sometimes. Like even if I want to have a massive production based game once I have built the third Industrial Zone I really don't want to build the fourth. At least not until I have 2 Theatre Squares, 2 Campus, 3 Commercial Hub, 1 Encampment, 2 Harbour, and 1 Aqueduct to hit all the Eurekas. The Eurekas create these very odd breakpoints in utility that feel out of line with the rest of the game.
This also pushes the game towards a lot of sameness. That isn't necessarily an issue, because having some incentive towards a balanced approach isn't a problem. It is perfectly fine to have some game elements rewarding players who build a bit of everything so that extreme strategies aren't quite so powerful. Unfortunately if you don't have a ton of cities to do different things in you are generally going to want to build your civ the same way a lot of the time, or at least with a really standard sort of core.
I was thinking about how the system could work to accomplish similar things without the same level of sameness in every game. For example, if instead of having a Eureka based on having 2 Banks, what if each Bank gave you a 50% chance to get the Eureka? If you structured all of the Eurekas this way you could set it up so that each time you fail a Eureka your next Eureka attempt of any sort goes up to 60%, stacking up 10% each time you fail. When you succeed, it drops 10% instead. This ensures that over time you will get the right amount of Eurekas, but you won't be doing the same thing every game. Overall Eurekas should stay about the same level but you will have a lot more differences in each game, and you won't have the same weird breakpoints. Three Industrial Zones to go for the Workshop Eureka is fine, but four is a little better yet as it gives you more opportunities to get it, and two is also ok.
Obviously that isn't going to happen, but I like it better as a system to keep the game fresh and new each time.