Recently, Courtney asked on Google+ about treasure generation. I left the following comment.
Like Jeff, I let players roll on the official tables (Labyrinth Lord) if they went somewhere I didn’t prepare and defeated or outwitted a significant foe (in my last game, it was a black dragon).
When I prepare treasure, I simply roll on the tables. The mechanized aspect of this task encourages me to believe in the independent existence of the fantasy world. I use surprising results to think of new twists: dragons without treasure got robbed, a bunch of hobgoblins with immense treasure are on their way to pay taxes to their overlord, things that enrich the world.
I never embellish monetary treasure. That includes gems and most jewelry. It’s simply there to be brought back to safety and to be converted to xp when spent in some way. I might embellish jewelry if it is expensive (500gp or more) and say that it’s a diadem, a crown, a bracelet, a holy symbol of a named person, an aristocratic helmet of a dead people.
When it comes to magic items, I still roll on the tables, but I try to embellish every single one of them. Potions have colors and taste according to a big table I keep adding to as the game progresses. Scrolls are spells from a named wizard, are psalms in praise of a particular god. +1 swords are elven swords; better weapons will pour cold, green mist when drawn, will have runes identifying their former owner (and if possible I’ll then add the same runes to magic armor to be found nearby). Again, things that enrich the world.
I find there’s also a problem of delivery. How interested are your players in actual detail if it isn’t actionable? If it is: How many pieces of actionable treasure can they take before being overwhelmed? Three? Five? Seven? How interesting will they deem the follow-up quests in comparison to delving deeper or exploring further or pursuing their existing goals? That’s why I prefer bland monetary treasure, personally.