Ummm, so much wrong, a Gish Gallop of wrong repeatedly wronged. I congratulate you, I've not seen so much sequential nuclear tech wrong since I accidentally listened to a couple of minutes of Helen Caldicott on a Youtube video.
No-one has built a LFTR ever so saying definitively what they can do is a bit of a stretch. There's a lot of Youtube videos, Powerpoint presentations and grad student TED Talks about them but no actual operational experience. The molten-salt reactor at Oak Ridge National Laboratories failed in many ways and it never ran with thorium (it used uranium). "Thorium" fuelled reactors actually burn uranium, the thorium is bred up into U-233 before being fissioned to produce energy and radioactive waste. Thorium by itself is not fissionable.
Uranium is commonplace, not rare. It's simple to refine. Sources are not limited. It is incredibly cheap to buy on the open world market and a number of large mineable deposits are not being exploited because the market is effectively saturated. Uranium as a metal is not particularly toxic, in part because the common forms are very insoluble and can't enter the bloodstream or digestive system easily. The damage, if it does cause any, is limited to some effects on the renal system. It's not evil poisonous stuff like beryllium or cadmium or arsenic or lead or a dozen other elemental toxins with low LD50 and bad neurochemical effects.
No nation extracts plutonium from civil nuclear reactor fuel -- it's hopelessly contaminated with Pu-240. All of the nations with nuclear weapons made their stocks of nearly-pure Pu-239 is specialised reactors in places like Windscale and Hanford. All of those nations have more weapons-grade Pu than they will ever need, indeed it costs them to securely store the surplus after the number of weapons held was vastly reduced in the 1960s (Britain had over five hundred nukes at one time, today it has about 140 or so). Any power reactor built after the early 1970s had nothing to do with making extra nuclear material for weapons. There were a couple of designs that could be co-opted to make weapons-grade material, the British Magnox and the Russian RMBK-4 but I don't think they ever actually did since there were proper weapons-grade breeder reactors available to both nations.
 It's possible North Korea tried reprocessing spent commercial reactor fuel in the early days of its nuclear weapons development. The first few tests did not work out for them, you may recall.