Reversible transactions are interesting but I don't follow how it can lead to the conclusions you mention below. I've always known reversible transactions were possible but that seems to be about disaster recovery which I admit is good for resilience and critical for security. As far as using arbitrators, and dispute resolution, this can be done but Tezos, Tauchain, and Ethereum can all do this so this isn't a technically EOS specific feature unless you are saying EOS can somehow do it better? As far as I now maybe EOS is the first to do it but once everyone else sees how it is done they'll just do it on their.

All current crypto's have the problem that holders collectively have to pay the price if a criminal manages to steal coins. I don't know how high Bitcoin or Ethereum would be by now if there never was a Gox or DAO situation.

This is very true. Ethereum is insecure. On the other hand Tezos is very secure at least from the specification, and as flexible as Ethereum. In a sense it is like Ethereum done right in terms of a ground up secure design. Tauchain theoretically from what I know has some surprise features it is capable of which none of the others, and Tauchain will have the benefit of learning from the mistakes of the projects which came before it, just as EOS does not duplicate the mistakes of Ethereum, and Tezos does not duplicate the mistakes of Ethereum.

So I guess my point is what EOS will have going for it is network effect and community, not technical specs. Tezos and Tauchain will be able to equal or perhaps even surpass it in technical specs but that doesn't mean it will have Steemit community, Bitshares community, Peerplays community, really next to Bitcoin and Ethereum this is the biggest most active community. Also if EOS smart contracts are a familiar looking language then I'll be able to program it right away, while with Tezos I'll describe it as an alien programming language, not because it's hard but because it's a different ecosystem with different development style etc.

As far as costs you have a point but I think cost of development is the flaw Tezos could have which EOS likely wont have. If you can spend less to get the same quality code then this is more bang for your buck. I have to see more of EOS or at least a whitepaper before I can really respond in a way which is coherent so don't take my criticism too seriously.