I regard drivechains as one of the most interesting concepts to scale Bitcoin. The problem is that there is an opcode (= a command in Bitcoin's script language) missing for it to be implemented.
In all sidechain concepts, the "hard problem" is how to guarantee to return Bitcoins to people who want to "return" to the main chain leaving the side chain. Basically, in the proposal, miners are in charge of this task. It is, however, a process that takes a large amount of blocks (and time) to make it secure against attacks.
That could be seen as a disadvantage. But if Bitcoin implements Segwit or another malleability fix then it will be possible to trade sidechain tokens to main chain Bitcoins via atomic cross-chain trading. The value of a sidechain token, if traded in that way, will be surely a little bit lower than a main chain BTC but I don't think the difference will be more than 1-2% because if you have time you have the option to return to the main chain via the drivechain mechanism at a rate of 1:1.
its more nonsense "Bitcoin as a settlement network" stuff.
I consider the possibility to scam someone using a strong sidechain/drivechain far lower than in proposals like the Lightning Network. So the need for on-chain settlements will be most probably much lower.
