In FE, Gold and production are closely related, as gold is more or less buying the same things production does, except for army upkeep :

For instance, items have a production cost when training troops, but a gold cost when upgrading troops or decking out champions. When rush buying, to directly turns gold into production. One of the current problems is that getting gildar from taxations has a way too severe penalty, which makes adventuring the only reasonable way to acquire gold.

When the item shop will be toned now so that the prices are not completely ridiculous anymore, we will still have a big economic problem that won't go away by itself, mainly that getting gildar will be too punishing, so no city will ever be seriously garrisoned, and more importantly, taxation will still be something you should really set as close to zero as possible. So there still won't be any choice like military expansion or economic development : everyon will just try to optimize its army so that they have the smallest army that can be good enough to clear the monsters in reach.

I think most production vs gold balance problems would go away if they were obtained the same way, as in Master of Orion 2 (for those who missed this great game, or don't remember, taxes in MOO2 directly cut the amount of mineral production you produced, but the amount you got was proportionnal to the tax rate).

I think it would work even better in FE, because of the cost duality of most things. That would make upgrading cost and item cost easy to rebalance because the prices won't appear so schizophrenic.

 For instance, there would be a fixed conversion rate between Gildar and Production, and basically, gold would just be stored production, and a 50% taxation would only mean that 50% of your production would go into Gildar (and not cut down on research (which I think would really help as the current system is way too punishing).

If the rate is 1:1, and item needing 10 production point would cost exactly 10 gold to acquire (or to upgrade troops), and would sell for 5 golds, not 1250 gold to acquire like it is the case now.  This way, questing costs/rewards and production costs would be automatically balanced (we don't have to make the shop sell items at a 1:1 ratio, though, it could still sell them for more, but certainly not buy them for more than their base cose).

Rush buying should still be more expensive, something like 4 times for the 50% first parts of the item production cost, and 2 times for the 50% last (or anything like that, these are just exemples of the variables I think the system should use). 

Trade goods should be an item you can produce (as in MoM and MOO2) : you would be trading current production for future one. We just need more things to produce from the beginning so that 90% of the cities do not produce trade goods, instead of not allowing trade goods because otherwise, every city would just be pumping these up (because currently, 95% of the cities produce nothing 95% of the time, because everything needs a tech unlock, which I think is another big problem in the game, but a problem best left to be discussed in other threads...).

Buildings that gives a % gold increase would instead just increase the gold generated by other building (once again, like in MOO2, at least, like I recall it worked), and give a fixed amount (ie they would not give this % increase when turning the production into gold, either by taxation or by producing trade goods).

The problem with this system  would be that taxation would not affect unrest at all. So what should affect unrest? 

I don't really have any idea about that, but if unrest is to stay, it could maybe only happen in conquered cities (although that would make people go on a razing spree and commit klackon genocides as in MoM), that would need a gallow and other such buildings to operate correctly, or it could affect cities that are too far away from your capital compared to your prestige :

for instance, if you have 4 prestiges, you could have cities up to (A+4 )*B tiles away from your capital without penalty (for the local city only), and up to (A+4)*2B away from your capital with a moderate penalty, like it worked in Civilization 1. With A and B being parameters that should be fixed depending of the size you envision for player empires.

Or just an empire wide penalty when the number of cities is too big compared to your prestige. 

I may lack inspiration concerning what to do with unrest, but it is mainly because I don't feel it work that well in this situation : "well, the realm is under constant attacks by horrible monsters, and we have everything to rediscover, the only shield between us and the threatening world around is our sovereign and his armies, but let's just get on strike because we don't want him to pay these lazy bastards". That just does not work in an epic setting at all imo. That is why I think unrest should be either tied to the origin of the city (ie founded by the sovereign), or the faction prestige.

That's why I was looking for ways to make unrest be relevant for something else completely, which would be to slow down the player after he has conquered new cities, so that they cannot reliably pump troops right after being conquered. 

These changes may be a bit radical, but I don't think the schizophrenic nature of the system can go away as long as we have 2 completely unrelated ways to produce things: through gildars that only depend on population size and taxation level and adventuring loot, or through production that depends on a completely different set of variables. I think the costs will always remain too hard to balance with the current system.