I have no real knowledge of economics ... but I think I find it hard to imagine a idealised market (= optimised?) for products with utility for so few people - with the usual meaning of utility.
Equally, I guess there are markets that operate for extreme vintage wine and obscure roleplaying books. What happens in standard theory when a small group of people want rare items? I guess everything hinges on the efficiency of those brokers - or for the RW, now, Ebay.
There are two levels of market in play here: the first is the market of the setting (the utility to characters), and the second is the market of game itself (utility to players). We pretty much need to make sure the relative valuation of items works well when considered as a game, otherwise the game doesn't work. It is my opinion that the valuation of items that is optimal for the game as game is basically the same valuation that would exists in an idealized market within the setting, since rational characters with the means and opportunity to optimize their equipment would do so, just as would rational players. (This doesn't impugn people who don't optimize their characters for DPS, it is just the fundamental assumption of modern economics, and very useful even when only approximate.) Since the game should work in a variety of settings, I think this is the appropriate starting point.
I'm not suggesting that we should assume our settings actually have idealized markets. Rather, the game should provide a simple rule of thumb to simulate the non-ideal factors. For example, in Dark Sun selling prices would generally be low, selling prices high, both subject to large fluctuations, and a good many items would be neither sellable nor purchasable because there is neither cash nor an equivalent item available. In Forgotten Realms these issues might be less severe, and even the most expensive items can eventually find a buyer. In Eberron many inexpensive magic items might be about as liquid as cash, but more expensive ones unavailable.
You could even extend these ideas to factions, individuals, and different geographic regions in the same setting. For example, there might be heavily regulated and/or discouraged trade of magic items in some city, so that the official channels are very illiquid. There is also a thriving black market, where you can get better money. If the characters don't think they'll just make up for the difference in their next treasure packet because the rules say they need so much money, suddenly breaking the law might seem pretty tempting.
Let me give a specific example off the top of my head that might work in 4e. Suppose we defined various levels of liquidity in markets by base selling price compared to canonical value and (if desired) price fluctuations for both buying and selling.
Code:
Liquidity Selling (%) Fluctuations (%)
--------- ----------- ----------------
Negligible 0 25
Low 20 20
Moderate 40 15
High 60 10
Very High 80 5
Complete 100 0
Then a faction would list a liquidity and an item level. For each additional set of levels above that given the liquidity drops by 1. When it drops below Negligible items above that level are totally unavailable.
So, perhaps a tiny village might be (Moderate, 1). It would have low liquidity for level 2 items, negligible for level 3 items, and level 4 items are straight up unable to be moved.
Sigil might be (Very high, 5). In that case, it drops to negligible at level 25. The City of Brass, the height of multiversal commerce, might be even higher.
Such a scheme would also enable some interesting haggling options, say tilting the fluctuation in the direction you want, or be the basis for rules regarding a trading caravan. The uninterested DM can just pick a level and stick with it. Regardless, if wealth by level is removed from the rules, the game should be able to tolerate any of these cases seamlessly.