There are a few games with reasonable economics: Runequest (2nd or 3rd ed, not the Mongoose versions) and Pendragon (all editions). Fantasy Wargaming, for all its derision as a game, has decent econ research. Later versions of Chivlary & Sorcery also do reasonably well at it. Several supplements for Hero System also have decent price lists.
There are several using abstracted resources: HeroQuest, Burning Wheel, Mouse Guard, d6 Fantasy, Rogue Trader, Deathwatch.
The problems with reasonable economics:
- Barter
- Unrecognizable units
- Localization
- Lack of data
- Playability
Barter:
until the 1700's, 90% of an economy was in simple barter; that peasant wouldn't take your coin of gold because he would be called a thief for simply having it. He might be able to handle a denarii or two, or the local copper...
Unrecognizable Units: Groats, florins, farthings?
People complained so much about the old "LSD Money" that the UK did away with it. And it had more than just pounds, shillings, and pence, but we know it for those three. (The English denarius, inherited from Rome via the French denier, became the penny by devalation, and the abbreviation for pence is "d"...)
Farthings, ha'pennies, florins, crowns and halfcrowns,sovereigns (20-22 shillings), Guinea, groats... and gold coin measured by weight and converted to silver at going rates!
Simply put, the variety in England alone was so diverse that it needs a score sheet, and most coin was copper, silver or gold. It can be shown, however, that the peasants seldom used coins until the industrial revolution.
Localization:
Each economy had a different set of valuations. A billy goat might be worth 5 chickens in one community, and 20 in another... depending on how well the goats and chickens do in those areas. Not to mention each king keeping different standards of measure.
Lack of Data:
while there is a wide amount of mercantile data surviving, it's not a uniform mix, and it's mostly bulk trade data. Finding out how much a particular item cost in a given year is a matter of approximation in most cases. We do have some surviving 1st person accounts in letters, and more in court records, but very few are consistent lists of goods, let alone entire systems. And where we do have lists, they often are the same base items at different relative valuations. So most of the extant gaming resources are derived from these same few extraordinary source lists, which were usually written about at the time for having been abnormal in the first place, hence the court cases.
Playability:
Playability requires some simplifications. The biggest ones are universal measures, universal coin sizes, and unified price lists. With only those, for the sake of GM sanity, one can readily research historical prices, and come up with a workable system.
Which is exactly what Greg Stafford did. He did this for both Pendragon and Runequest.
Likewise, he examined historical documents for knights' manors and fees... and the disturbing thing is that the divide between rich and poor is low.
Adding magic can and should alter things, but, fundamentally, we can't know exactly how much. For example, in Tunnels and Trolls, a wizard can comfortably retire and break whole guilds with one spell: Slush Yuck. He can, for a short period, turn stone to slush, and it can be put in molds, and when the spell ends, custom cut stone. Which means, long term, custom stone homes being unreasonably cheap... successful peasants kind of cheap. And mountains with whole communities cut in by a single wizard for the fun of it.