I love what you're doing. Philosophically, I'm behind it 100%. The problem: all that most customers really care about is price. They're willing to eschew all other concerns (e.g., security) if they can shave a few pennies. At the end of the day, your quest may end up being quixotic. And that's a tragedy, because local-first should have broader adoption. :(

Look - I love good, local software. I hate paying for webshit.

That being said, you are going to struggle to compete if your main selling-point is local operation. If you're trying to target the developer audience by building UNIX-like tooling for your software service, you're eschewing a huge market to chase your least-likely customers. You will be replaced with a free version, if not end up with your tool being forked or pirated in it's entirety.

SaaS is the way it is for a reason. It's very quick to iterate on simple ideas until they're profitable, as opposed to solving hard problems and hoping people will pay for them. Just because we've ended up in a sea of webshit doesn't mean the solution is paid coreutils though - most people evaluating your product simply won't give you the time of day.