This actually has a similar purpose to an idea I had for basically creating a "your own backoffice" distribution of software (a little more turnkey, but with the same idea of tying it to a particular low-cost cloud provider). I love the concept and love seeing this kind of thing. I believe in the mission. I think being opinionated about first steps is an important part of it.

What bogged me down were "day 2" concerns around troubleshooting and maintainability. I had a somewhat more ambitious plan that would involve monitoring and rollouts and so forth as well, so it's not really something you do in a shell script and forget about. And I was focusing on the use case of a small group of people rather than a single, disconnected person, because that's so much of what (I think) we use our devices for. And I wanted to focus on using tools in a standard way so that you could actually learn how your system worked by looking at its standard configurations, written in a standard way (not "pkg_add" and "opaque ad-hoc config template") and being aware of how to change them over time as you learned more.

I still think this is great but I'm not sure it's the kind of thing you can just run and then be okay. Hell, I'm a "tech professional" for whatever that's worth and it's hard enough for me to keep homeassistant updated and free from rot. A single server that's not backed up--and even if it were, without features for restoring--or updated (even recommending never updating it--I love OpenBSD and its stability is great but there are reasons to update) just isn't something to tell someone to start relying on for their digital life.