Good article.  

> So the main thing that Satoshi got around the problem of trust, is to create a system, where nobody needs to trust anybody in the system as the system creates the trust by default

But, do end users (not software developers) of the system need to actually trust the software code release that makes trustless consensus operate?  

The system in operation (software, network of nodes, etc.) does indeed create a "trustless" environment between parties, assuming enough parties continue to participate to keep it sufficiently decentralized.  But as an end user (assuming no inspection of software code is realistic), I feel I do need to trust the open source software that runs the system.