Celebrim, I'd love to read your thoughts on this. Even a quick summary / pointer of what sorts of things you have in mind would be great!
I want to write a book, and you want "quick pointers". I'll try.
1) All actors in a world need to be using the same economy. That is, a 'days wage' for all actors in the economy ought to be similar and based on the same coin. It makes absolutely no sense for a 0th level fighter to want 3 s.p. per day, and a 1st level fighter to expect 100 g.p. per day in wages. While it doesn't at first seem to be a problem that peasants and non-skilled laborers (those with jobs that aren't immediately helpful to a dungeoneering PC) use a silver piece economy, while PC's use a gold piece economy, what you discover is that it becomes way too easy for PC's to leverage the silver piece economy to for example - buy all the unskilled labor in an area, essentially bribe all the NPCs they meet, or otherwise gain a monopoly on production. Rationalizing the situation only makes it worse. For example, if the noble NPC rivals of the PC's have the wealth to compete in that market, what that means is that all the peasants see that the PC's are munificent and charitable, while their erstwhile lords are greedy selfish hoarders of wealth that have unfairly left them in poverty. One simple fix of this that has always worked is convert all base prices of everything to silver pieces and downgrade the size of treasure. You could successfully do that even back in 1e, just by anywhere you say 'g.p.' reading 's.p.' and dividing the gold and platinum columns of the treasure types by 20. One huge advantage of that to everyone in the economy, is it no longer requires a caravan or camels or mules to carry enough gold to market to make a major purchase. I remember the first game I ever DMed, the players ended up pulling away from the dungeon with a wagon carrying 30,000+ g.p. - 1500 pounds of gold! Given the wealth = XP assumptions that was basically inevitable if the party was to be 4th level or so after delving to the bottom of a four level dungeon, but even with later game assumptions about the cost of items that's still not unlikely.
2) The baseline economy needs to be an ordinary economy not an extraordinary one. If you read the 1e player's handbook closely, it will become immediately obvious as you read the section where Gygax discusses the price list that he's not actually describing a baseline economy. Gygax was not historically illiterate. He well knew that the prices he was giving for most of the dungeoneering equipment was gamist and designed with balance in mind, not historical accuracy. But his simulationist inclinations lead him to seek an in game justification for his game rules, and what he hit upon was that the game was taking place in a setting equivalent to the Klondike gold rush. Lots and lots of gold was being hauled out of these ancient dungeons by intrepid adventurers, and everyone was rushing to get in on the chance of wealth. As a result, the cost of adventuring goods was experiencing hyper-inflation, as demand outstripped supply. People would still pay the prices, because delving in the dungeons was so potentially lucrative, but these were not normal prices for swords and torches and what not. Unfortunately, that critical fact was seldom acted on (even by Gygax), and even if it was, it answered nothing. I don't know how many published modules I read that said, "Prices in town are double/triple of those in the Player's Handbook, owing to the scarcity." Double or triple the prices in the Klondike during the gold rush? Really? More to the point, if these prices were not realistic prices in long settled lands without hyperinflation, what were realistic prices? And if there is so much money to be made selling this hyper-inflated goods to fools, why don't the PC's try their hand at that rather than going into dungeons? Surprisingly, through all the editions, Gygax's price list has never really been challenged. It's been tweaked a bit, but the core ideas still influence prices in modern editions. And really, that probably extends to a wide range of RPGs that are influenced by D&D. What we really want is a transportable economy, that we can tweak to fit a setting no matter what the GM's assumptions about the wealth of the setting happen to be.
3) Two thirds of the cost of everything is labor. If someone crafts something, the added value should be proportional to their wages/salary. Conversely, you ought to be able to take a price and work back to the salary. If an item is the product of skilled labor, and costs 15 'days wages', then you ought to be able to work back and say something like, "The raw materials cost about 5 days wages, and a skilled laborer spent 2-3 days working the raw materials." If the item is the product of unskilled labor, you ought to be able to work back and say something like, "The raw materials cost about 5 days wages, and a skilled laborer spent 5-10 days doing the work." The raw materials in turn are probably the product of some unskilled labor doing 1 day of work, and a skilled laborer spending a day finishing the work, and some amount of cost transporting the good to you. If those estimates don't make sense, it's likely that the prices don't make sense either. As a practical matter though, working the whole economy from those first principles is going to be hard. It's just a useful check, and actually, it's more likely you have a solid understanding of the final cost than you do the process used to produce it. So its more likely you'll be able to use it to answer questions like, "How long does it take to make a sword?" or "How long does it take to make a wardrobe?" You may not know, and if you can't immediately find a good answer, working back does a fairly good job. Ideally, the crafting system generates this working the other direction, so that a player with X skill in making stuff can produce value added goods of the right prices.
4) Luxury goods are a partial exception. If you make a wardrobe out of cheap wood, it takes about the same time as a wardrobe made of rare and expensive wood. You can add some additional cost based on the risk of ruining the rare material and the fact that someone who only serves the luxury market has less consistent revenue (fewer customers) and so has to make more profit on a few sales, and there might be more adornment (art for arts sake) but the more valuable the luxury good the more and more of the cost is in the premium placed on the raw materials.
5) Historical prices are hard to find, but historical costs of labor are relatively stable over time. This is probably the most powerful technique I've ever discovered. Manufacturing and industrialization have made certain types of goods much cheaper (in terms of days wages) than they ever were before because they don't take as much labor to make, but the cost of hand made goods is actually reasonably stable over time. If you price a hand tailored set of clothes or a hand made piece of furniture or a hand made piece of fight quality armor or even a house, you're going to end up with a price that is reasonably close to the historical price of the same or equivalent good. The distorting factors like the reduced cost of raw materials owing to industrialization is going to be no worse than the error introduced by being unable to correct for local factors or inflation when inspecting historical bills of sale. This gives you an incredibly powerful tool for correcting and sanity checking price lists, because it gives you a conversion rate between historical and fantasy prices, and modern prices. You can actually reasonably make a translation like, "$50 = 1 s.p." or "$1000 = 1 g.p.", provided you aren't using primarily machine made goods as inputs. I didn't really have access to the full power of this technique until selling things on the internet became ubiquitous, but the very first example that pointed me in this direction was gems and jewelry. It was obvious even to my 10 year old mind that Gygax's gems and jewelry were gamist artefacts intended to be transportable XP for high level characters and not realistically priced goods. For example, one of the cheapest pieces of jewelry you could find was worth 200 g.p. But that is in fact 4000 days wages. If you plug that in to any of the sanity checks I've described, you'll get nonsense. If you use the price stability sanity check, that means that you'd expect silver jewelry on Etsy to be on the scale of $200,000 for each piece. If you run the two thirds of anything is labor sanity check, that means that the raw material for the jewelry was about 60 pounds of raw silver and that a skilled laborer spent a year and a half working the silver. I don't know about you, but a 60 pound ring or bracelet sounds insane. And in terms of everyone being on the same economy, this means that even the simplest piece of silver jewelry is well beyond the peasant's means, even though historically women prized jewelry because it was portable wealth that they could own even in an economy where they weren't allowed equal access to property ownership. So we know historically, even the poorest peasant women tried to develop a 'retirement fund' of simple jewelry they could pawn off in an emergency. Using a sanity check, simple silver jewelry ought to often be worth no more than a day or two's wages - which agrees with prices on etsy and historical prices. Even gems or jewelry that we consider quite valuable, $3000 or $30000 dollars, is only worth a handful of gold pieces or a few score of days wages. However using the historical prices are stable, we can also work from the modern costs of rare and expensive jewelry to what the crown jewels are actually worth, if we ever want to fill a chest with the royal handiwork of some lost dwarven kingdom.
6) As much as possible, the game shouldn't assume the wealth of the setting or the characters in it. An ideal discussion of economics in the game should be able to 'good enough' cover everything from a D&D game set in the stone age, to a D&D game set in the early modern era just before steam power and industrialization really starts to make the economy alien to someone a few centuries before. Heck, even that could be overcome if the era when a good becomes industrialized is flagged, but that's probably more detail than any system has ever attempted (and as a practical matter, may ever need).
7) As much as possible, the game shouldn't assume what game is being played at the table. Where this shows up most powerfully in D&D is the level Gygax assigned to spells, and the cost he assigned to magical items. Gygax assumed that the core of the game was PC's delving in dungeons, and he made his costs and accessibility of goods based around the question, "What's useful to PC's in a dungeon?" And that's fine as long as your game stays in the dungeon and has minimal interaction with the world outside of the dungeon/haven system he assumes, but even in the 1e DMG, it's obvious that Gygax's simplified world doesn't reflect how the campaign and the player's interests are going to evolve. Gygax even recognizes and encourages it, but he never really corrects his gamist pricing assumptions to reflect the changed game he's playing. So, the cost of his lairs perhaps at a stretch reflects the cost of building a PC's prestige stronghold in a wilderness area with little skilled labor and no preexisting infrastructure, but tells us little to nothing directly about the cost of peasant housing or building that infrastructure. This has all sorts of subtle impacts. For example, Gygax prices a 'mount' spell - get a really great horse for a day - in terms of its utility to the player. Having a horse is a nice perk for an adventurer, but not that game altering in terms of a game built around delving into dungeons. But the spell is game altering in its economic impact. If a really good horse costs 60 days wages, and requires 2 days wages per day in upkeep costs, then the value of a mount spell to the game world is incredible. It's pretty obvious that society would quickly try to capitalize on this value, by having messenger, courier, or taxi services offering this incredibly valuable service for the very low cost of a 1st level spell. Any 1st level wizard now has immediate employability and huge return on investment. The medievals developed the university system to produce high quality civil servants, how much more so would governments try to develop and capitalize on spell users! This is cultural and game changing, and you can't ignore it, because as soon as the PC's themselves get in a position to capitalize on that economic returns, you can bet they'll set up their personal kingdoms to do so. So if you don't want a world overrun with Magical Pony Express, you have to adjust the costs of the spell to make sure it doesn't outcompete ordinary horses and change the trappings of your world. Consider then the implications of something like a Lyre of Building in terms of its return on investment!
Personally, I'd pay $69.95 for a book that actually did this right. I might pay even more if it had a really good abstract system for running dynastic play at different levels of granularity - ei, how much income does a PC get from the properties he owns. Pathfinder's Complete Campaign had some pretty cool rules along those lines, and I've seen bits and pieces here and there that were really cool, but I've never seen a designer really put all of this together.
