If you can't bother to understand the work then heuristics are your best bet. In this case, the authors are leading academics from good computer science departm

If you can't bother to understand the work then heuristics are your best bet. In this case, the authors are leading academics from good computer science departments. This makes the probability of the paper being an outright hoax very low. They would be risking their careers and the respect of their peers - and for what exactly? Shamir in particular is already at the top of the game (full professor + turing award + many millions of dollars from royalties). Now, mistakes CAN happen. But in this case, it would be very hard to make one. If they built a system that reconstructs the cryptographic key then you could compare the keys on both sides.