That assessment of "hundreds to thousands" of years of course takes inventing breeder nukes. Since the natural fissionable uranium is 235 at .7 of a percent, we need to efficiently make the U-238 into Pu-239 to use. Also, with breeders, we can use thorium by making it into U-233, an artificial fissionable. You need to isolate U-235 (already done) to get the ball rolling once you invent a breeder.

Of course, you don't want idiots to get any of the fissionables, as they can get a bang out of them. This is the biggest problem of nukes in general. Breeders mean that you can recycle waste most-way. Spent fuel is mostly U-238 with some Pu-239 added for flavour from normal reactors. The Pu-239 comes from neutron bombardment of U-238 during use. A "perfect" breeder would end up making waste made of fission byproduct elements about half the atomic weight of the fissionable.

As above, fission has its problems, but nothing is problem-free. I'd rather risk an occasional Chernobyl vs. the certainty of global warming with using up the coal. With fusion, all bets are off. It could take centuries to figure it out - assuming we keep civilisation up and running! (and that's a dangerous assumption!)

Of course, you don't want idiots to get any of the fissionables, as they can get a bang out of them

Of course, many would argue that many of the world's leading idiots already have fissionables.

Unfortunately, ONE idiot does have fissionables, in the form of the remote to use the fissionables. That is guarenteed, until 1/20/2009. (at 11:59:59AM DC time) Barring the possibility of this idiot seizing power in a coup-like action.
One alternative is a slightly improved CANDU reactor (using heavy water) with a lot of thorium mixed in between the convential fuel.  1:1 breeding (no extra fuel for other reactors like French Phoenix, etc.) but no real depletion either.

From my Physics background (decades ago), U-233 is not an ideal bomb ingredient but can be made to work with an effort.

Bush has decided to push MOX recycling (do not seperate all of the relatively short lived and intensely radioactive elements above Pu).  These elements have some fuel potential as well.  The heat and radiation from these make fabrication from stolen fuel "problematic".

Estimates of fuel reserves are with current technology & prices.  Uranium has been prospected for and mined for only a few decades; much less than most other minerals.  No interest in new sources for two decades or so.

Fission reactors have considerable prospects for that "New Technology Silver Bullet", unlike oil.

BTW: Used fuel may become a good source of platinum group elements (when U atoms split, they do so in a variety of ways).