Follow-up question: let's assume one of the signers can no longer be trusted, or disappeared (this will eventually happen): how do the other signers decide when to create a new multisig address with a new signer? Is there any centralisation involved?