There lies the inherent problem with "trusted setup", why leave there a possibility of 6 people not colluding? That just isn't scientific or exhuastive. Just drop the IFs, go with Monero's method of RingCT.
Beside, Monero's privacy is a working feature today, with no pre-mines, unlike ZCash.
Essentially they each produced a private key, and if each of them revealed their private key to the same party then that party could derive a master private key that would allow them to (among other things) mint Zcash for free (privacy wouldn't be broken). Assuming that any one of the 6 did in fact destroy/corrupt his private key without revealing it, then the collusion opportunity is forever lost.
"Until the software and deterministic builds are audited, the entire ceremony is a bunch of crypto hocus pocus that means nothing."
The trusted setup is really dubious and highly vulnerable; so far my efforts were wasted due to the fundamental problems that everyone in the multi-party-computation ran the exact same unaudited software.
The trusted information (supposedly) can't be given up now because nobody has it: it was ceremonially destroyed in interesting ways.
1. All people present in the trusted setup would have to be colluding or be compromised to cause an issue.
2. Some of these people have reason to make the currency succeed, because they own large amounts of zcash.
3. Some show the lengths they went to to prevent compromise: https://petertodd.org/2016/cypherpunk-desert-bus-zcash-trust...
One of my own opinions on collusion:
If collusion was suggested by one or more of the people present in the setup it would be quite likely that another one would have come out and said something about it. This would have made zcash and the person itself very untrustworthy. This makes it likely that it wasn't worth the risk to the person thinking of collusion in the first place.