Actually, in 0.951, I've really started to see the impact of population in my games. No longer is it a 'bottomless void' to be filled. The growth has been capped really well. In addition, the additional population you receive from combats accelerates your cities expansion and further caps your population. you need to consider growth strategy now.

Taxes definately need a tweak though. Having them directly tied to production isn't the way to deal, imo. I'm fine with that being the end penalty of unrest, but there should be incrementing penalty per turn for too high of taxes. There should also be incremental benefit per turn for too low taxes. Thus, if you wanted or needed to run your country a certain way, there would be trailing consequences.

For taxes too low: maybe improvements 'downgrading' or roads falling apart, or population growth too quick (which can cause a disease event to occur), etc

For taxes too high: increasing unrest, workshortages, outright rebellion, population decrease.

Taxes should also not be hit/miss. Each level of tax rate should have pros and cons; and it should be up to you as the player to weigh and consider what consequences you are wanting to deal with.

perhaps as Magnar, you like you taxes high, cause anyone who rebels...you just turn into slaves?
perhaps as  Tarth, if you have enough 'over population' you have a chance at producing a free pioneer?

Stuff like that...

It sounds like you are talking specifically about Magnar, with their ability to capture population.  Even at that, I don't see how capping growth affects anything other than your access to level up buildings, or how population accelerates your expansion.  Building buildings expands your cities regardless of the population.  That is the problem we want to address.

Nice discussion but I'm loving the way population is implemented. It's clear cut, it works and it's NOT overly complicated like so many other games try to achieve.

OK, Let me take another shot at this and see if this is simple enough and will work within the current design with only a few additions.

•     There are three types of people in your cities: Happy, Content and Unhappy.
•     Each city starts off with a base fixed number of content people (e.g. 80 people) and 0 Happy people.
•     The number of content people is reduced by a percentage based on tax level, so if you had low taxes, then the number of content people would be reduced by 20% (80 content - 20% = 64 content).  The number of Happy people is unaffected by taxation.
•     Unrest buildings add to the number of happy and/or content people. (e.g. +10 happy people, +80 content people)  Unrest buildings could also directly counteract the taxation penalty (e.g.  +5% content people) allowing for higher taxes without penalty.
•     All population over the total of happy and content people are Unhappy.
•     Only happy and content people work.  The total number of happy and content people is your productive population (ProdPop)  In order to make numbers that are based on ProdPop more readable, they are based on 100 ProdPop  (e.g. +0.05 research per ProdPop = +5 research per 100 ProdPop)
•     Everyone pays taxes, so taxes are based on total population (500 pop. * 5% tax rate = 25 gildar)
•     Buildings that generate gildar (markets, etc.) , production (workshops, etc.), or research (studies) do so in part as the result of the work of the population.  Their output is determined by a combination of fixed and population based bonuses.  (e.g.  +2 production, +2 Production / 100 ProdPop)
•     Fixed bonuses can be arbitrary or based on the resources of the city location  (e.g. +2 Production   or  +2 Production per Material, or +(Materials) production / 100ProdPop)

Most of the requirements for this to work is already in place.  All that would be required is the addition of variables for the max number of happy and content citizens in a city, and granting new cities a base value for content citizens.  The rest is just changes to the XML (which would not be insignificant).

What does this allow that we don't currently have?

1. Unrest increases as population increases above the Happy/Content threshold.  The current system has unrest as a fixed amount based on tax rate, and population does not figure into it at all.  A small village and a massive metropolis with no buildings have the same unrest.
2. The current system has the ability to base output on population, it just doesn't.  But it should.
3. Another added benefit is that you can have output that is affected by unrest (+2 production per 100ProdPop), and output that is not (+2 production).  The existing system lumps it all together and then reduces it by a percentage.  This allows you to moderate the effect of unrest better.
4. Happy people are a fixed number of productive population, content is variable.  This allows for a greater degree of fine tuning.  You could drop happy people altogether and just add more content, but then the effect of taxation would be more extreme.  If your pop. was all happy, it would be completely unaffected by taxation.  If it was all just content, it would be unaffected to the threshold point, and then the effect would be extreme.  A balance of both factors allows for a more moderate effect.
5. If Population affects production, then Pioneers can be made more expensive so their production early on is more significant.  Later, when you have larger cities with greater population, Pioneer production can be more commonplace.  In the current system, production that is based mostly on location materials does not increase much over the course of the game.  The relatively constant production requires Pioneers to have a lower cost to be available early on, encouraging spamming.
6. This has some interesting game play possibilities.  Like if your unhappy population exceeds X%, there is a chance per turn of parties of rioters spawning next to the city, destroying outposts and resources.

DexCisco is making a few good points, whether or not it's 'the' idea to pursue.

Fact is, with unrest linear, there is no incentive to queue buildings that decrease that unrest. If you are comfortable with a 22% production penalty...which actaully isn't that bad all considering, then there are a whole string of buildings that never need to be built.

If unrest increased with population, there would be a drive to consider buildings to reduce that unrest.

This is not true.  I am not suggesting having population be the only factor in output, but that it be at least included in the equation.  You could have buildings that add fixed production (+production), buildings that add by population (+production / ProdPop), buildings that add fixed production by materials (+production / material), buildings that add by population and materials (+(materials*X) production  / ProdPop), or a combination of all of the above.  This is not exchanging one variable for another, it is really just making an existing variable more useful.  Right now, materials are basically all that is important and that needs to change.

This is also not entirely true.  There is a limit to population right now, that being food, which is limited by +food/grain buildings and available grain in the location.  To say that you can't keep a soft limited number of people content with an unknown set of bonuses is a little rash.  If all of the available unrest buildings combined could keep at least 800 people happy or content (and late game buildings should be both expensive and very effective), then I don't see the problem.  There are also spells that permanently affect the unrest level.