You can generate a lot of gold this way.
Generally, I've found that I do better by continuing to expand. I sometimes have one city dedicated to failure gold, but it's a bit on an investment, since you can't predict when (or if) the AI will ever complete a wonder.
it's a little more interesting a technique to apply to national wonders - in particular, Hermitage, in a game where you don't really need to build it anywhere.
I would recommend you read Rhino's Failure Economy game for a good idea.
Yes you can only have something in 1 build queue at a time but the idea is to either slow-build the wonder in one city (preferably whilst owning the resource that speeds it up) or to be using whip overflow. Build a normal unit/building, whip it when the time is right and put the failwonder next in the queue. Next turn, unit/building is complete and the wonder gets the overflow hammers. Remove the wonder from the queue and repeat with the next city that's due a whipping.
T'is my understanding anyway.
Thx for the link. Read that this trick doesn't work in BTS anymore so I guess that settles it.
Still, thx to everyone for your replies
If you don't need the hermitage then you are wasting all the hammers in whichever city you do finally complete it in. National and Heroic epics are good because you are going to want these completed in a city eventually.