[MertsA]

The Uranium tetrafluoride itself is basically insoluble in water. But the fuel isn't really the issue, Uranium is a toxic heavy metal but in terms of irradiating the environment outside of a reactor it's not very radioactive at all. People always clamour over Uranium having a half life of billions of years but what that means is that Uranium basically doesn't decay on its own. It's fissile, so hit it with a neutron and you'll induce it to split, but once it leaves a reactor and is subcritical you don't have to worry about the radiation from Uranium.

What you do need to worry about is the radiation from the products of fission as those are the nasty components of nuclear waste. Not only do you have the direct products which have a tiny half life compared to Uranium, those direct products have a decay chain of their own and in most reactor designs out there those products stay in the reactor so not only do you have the decay to contend with, you also have additional fission since it's still being bombarded by neutrons. This is what leads to the toxic soup of radioactive waste, you get tons of isotopes of elements so while the fuel itself wouldn't be too terribly awful for the environment the partially burned fuel is dramatically more radioactive.

The fuel salt itself though if it's anything similar to what was used in the MSRE or FLiBe's new Thorium breeder reactor would actually be insoluble in water and would freeze, hopefully trapping the waste in the frozen salt just like vitrification of nuclear waste. I am not a chemist, and any knowledge I have on the subject is just the result of a hobbyist interest in the subject, but I would think that in the event of a MSR being sunk and breached that the frozen fuel salt would keep most all of the radioactive products dissolved in it safely contained.