I think the larger problem slowing down their growth is not his "neo-reactionary" thought bleeding into the architecture of Urbit. I think a big part of it is the massive technical debt that Yarvin straddled the project with via his C code. It is written in a hard to read style that I at least have never seen before anywhere else.
This is a pretty short critique about apparent issues with Urbit, from a non-ideological viewpoint and detached from referencing the author- though a reply points out those technical criticisms may have been inspired by ideology!
That is definitely a very insightful post you linked, but I think the project is terrible even if you block out every single possible connection between the tech and ideology. I quite literally mean that I can't read the code. They designed a variable and function naming convention which obscures what the C code is doing. The naming is all nonsense syllables and seems deliberately intended to confuse anyone trying to audit the code.
The app for example claims to be end-to-end encrypted peer to peer connections. Is this true? Is it implemented correctly? I'm not sure they could find a reputable security auditor who would willingly subject themselves to this codebase regardless of how much pay was offered.
[0] https://github.com/urbit/urbit/tree/master/pkg/urbit