For a random beacon, we need an honest minority assumption. Otherwise, using threshold cryptography as a random beacon, an attacker can very easily influence the outcome of the beacon.
That is why in the current design that @JustinDrake is considering, we are using VDFs. There is also an alternate design called the leaderless k-of-n beacon by @JustinDrake (Leaderless k-of-n random beacon) that looks very similar to the Dfinity random beacon that uses threshold cryptography. The issue with this design is that it would need quantum resistant cryptography in order to have the same properties as the Dfinity beacon.
The goal is to a solution that would work in the long-term. We are currently seeing quantum computers (even though they can’t do much yet) come into existence. So, we would still need to design a beacon that can last more than 20 years.