I feel like the article definitely has some merit but the timing of it (days before a TBD announcement on token trading), the inclusion of very obvious non-unique points (i.e. not scalable, anything Bancor can do Ethereum can do - uh, why not include the definition of Turing complete here) and the false assumption that Bancor's use case is for a groundbreaking ERC20 token that's going to be traded on every exchange makes the author's claim dubious.

A more likely use case for Bancor is some SMB (store/restaurant) or content creator issuing tokens to be used for membership reward/points or for voting on a new video/project idea, not for Augur/Steem/Gnosis 2.0 to issue an ICO.

If you take this gross assumption of the equation, his whole argument about slippage or "trailing the market", etc are gone. You're issuing some small time tokens. There no reason to always buy back from the market and there are no big players trying to "deplete your reserve." The author makes it sound like I'm selling some popular cryptocurrency at a fixed price and am being left to bleed dry when in actuality, I'm issuing my_random_token for friends and customers.

I didn't review the code but if what he's saying is correct, there is a reason for alarm regarding their reimplementation of basic arithmetic functions(!!!) and user overpaying.

By trying to make this into a laundry list, it comes more off as a FUD argument, very similar to what Bitcoin maximalists were doing post the DAO hack to Ethereum.

(Disclaimer: my Bancor holding is about 0.5% of my total Crypto portfolio).