For me, the problem with this coin is it effectively throws away the benefits of proof of work while retaining it's drawbacks, thus not functioning as a decentralized, permissionless system as cited in the first post. In that instance, you've created something that functionally falls into the same category as proof of stake, just with higher overhead and worse economic incentives. The economic incentives are so misaligned, the coin doesn't even work at all at release. The solution, CFB calls it, is "training wheels", or checkpoints.
The coin's value should theoretically increase as the security of the network does, at least that's how normal systems function. A baseball player that hits 100 home runs a year usually makes more than one that hits 0 home runs. With this coin, the security at release is 0, but we're supposed to believe the for-profit issuer of the coin that the price should be high when the network doesn't even work. We're then supposed to all willingly join in on this extortion scheme and pay for security now which doesn't exist, with the naive dream it will some day in the future. The way the coin is designed is not really conducive in creating a network effect for this to happen though. You're buying something that needs to go from 0 to large tipping point instantly with nothing in between. It's like hoping dogs will randomly begin walking on two feet and start talking out of nowhere.
Anyway, as I was saying, since the coin already functionally falls into the same category as proof of stake, and the outside entropy of the coin serves no real purpose, he may as well of just created a closed loop proof of stake derivative in the first place and attempted to scale via something like deterministic block production. You'd end up recreating some system like Bitshares, which like all proof of stake systems, is also a permissioned ledger, not anti-fragile, horrific fault/state recovery, etc etc. The point is, this coin is not a valid contendor or replacement to Bitcoin, neither are the proof of stake coins that share many of the same characteristics.
But proof of work is old. Its washed out. Get with the times
POW is far from washed out, it's just not being used in the most efficient way in Bitcoin et al. at present.
It is still, by far, the best way to regulate participation in consensus, and whoever manages to utilize it to its full potential, while solving the current inherent problems, will be sitting on a nice gold chair with a heavy crown on their head for years to come.
IOTA is one of a small few that takes a different approach to its use. Maybe it will catch on, maybe it will fall flat on its face, or perhaps even there will be a flaw, but at least someone is trying something new instead of the same old same old.