It's interesting that using a PoW blockchain such as Bitcoin as a random beacon is another way of implementing a pre-commitment. Since blocks are created upon another, and have to meet PoW difficulty requirements, miners are effectively pre-committing to which class of random value is the next beacon, while the economics (and provably difficulty of PoW), makes it expensive for them to manipulate future random values by discarding valid PoW solutions with undesired beacon values. Basically for every bit in the PoW beacon value that you want to fix, you have to throw away half of the potential block solutions even if you have 100% of hashpower: expensive!

Related: another clever technique that can build on any random beacon is to use a long hash chain to actually determine the final beacon value. Zcash actually used this for one of their trusted setups. Basically, they computed SHA256(x) iterated billions of times, a serial computation that took something like a week of solid computing, using a particular Bitcoin block hash as the initial value. There's no way miners could influence this, as there was no way that they could do the serial computation fast enough to know what actual beacon value they were creating.