Well, one explanation is that he thinks the threat model is such that a more likely risk is RCEs into the browser; if Tor directory authorities are doing untrus

Well, one explanation is that he thinks the threat model is such that a more likely risk is RCEs into the browser; if Tor directory authorities are doing untrustworthy things that can probably be more easily detected than deliberate exploits. It's also a very pragmatic plan: rebooting the whole Tor relay/exit node community is going to be a long term project without clear gains in the end; if your replacement looks like Tor in that you continue to invite volunteers to run nodes if the previous set of nodes are compromised what prevents those nodes from joining your network too? Arguably better to focus on forking what you can fix, which is apparently the browser. But that's all a charitable explanation: the less charitable explanation was he forked the browser because that's where the Tor branding is... Fork the relay codebase and you can't take pretty screenshots of your new fork.