Nice article! What do you think of the Bored Ape approach? They used a provenance hash at the outset to prove the sequence remained always unchanged, and then made that fairly trustless for their reveal by having a starting index function to offset the initial sequence by a random factor. I think something like this approach seems fair. Overall, this is still a tricky issue, and as a dev I find myself thinking about better approaches all the time, even if, on the whole, not many NFT *buyers* don't seem overly concerned (though they should be!). I predict a rock-solid standard will emerge at some point.