Safety is not about technology, it's about human error. As long as people do dumb things, no design will prevent a catastrophe.
Look at the three big reactor failures: Chernobyl, Three Mile Island, and Fukushima. All three were caused by human error. For Chernobyl, it was a dangerous design and running dangerous tests. For TMI, it was a less dangerous design, and they still screwed it up with bad procedures. For Fukushima, they made a series of globally bad design choices because they refused to consider realistic worst case external events. Plus they uncovered a flaw in the containment structure design that lead to the hydrogen explosions.
All of these are human error.
And it's not just reactors. The British Petroleum oil platform blowout in the Gulf of Mexico was human error. The sinking of the ferry Sewol in Korea was human error, as was the sinking of the Concordia off of Italy. BP also had a refinery blow up in Texas because of bad operations and ignoring a known problem with volatile fume leakage.
So no matter how secure a technology looks, it will still suffer a complete worst case failure. Assuming anything else is wishful thinking.
What's the worst case for LFTR? No one seems willing to even talk about it. It's remarkably like the head in the sand attitude that lead to the Fukshima disaster.
So here's a question: what happens when a molten salt containing fluorine, uranium, thorium and other miscellaneous radioactive elements comes in contact with water? Does it explode? Does it burn in air? How toxic are the substances entering the environment? (Trick question: both uranium and fluorine are very toxic elements. Fluorine forms many toxic compounds with carbon.) What is the equivilant explosive energy of tons of molten uranium salts?
If it is burning, how do you put it out? (Note: with fluorine compounds water is a bad idea. It's explosive.) How do you build a containment vessel that will withstand all of that? How will the cost of proper containment and emergency planning and equipment impact the economics of power generation?
A burning LFTR makes a burning graphite reactor seem like a campfire for a marshmallow roast. Good luck with that.