I got a question though (might be to the wrong person). No code is perfect, and we know that there might be attacks on the PoW-chain (seen in a lot of especially new chains over the years). What I don't understand with the whole "2 chains is better than one" is how the PoW-chain will keep tabs on the PoS-chain? Will everything fail if the chains aren't coming to the same conclusion? If I, in some way, manage to get a 51%-attack on the PoW-chain, will I just invalidate the block since PoS-chain isn't coming to the same conclusion?
It's not so simple. There are several mechanisms, and the pos is not a "controler" of the pow chain in the sens where it does not validate or sign pow blocks.
It's more loosely coupled than that, and merely observes to report strange behaviour from the pow actors.
Lots of partial docs are done and will need rework, check the official github, some reading are awaiting already
https://github.com/bismuthfoundation/hypernode