This reasoning obscures the huge overhead of creating a distinct "version 2" next year. I can do a much better job on the software if I don't have to spend 70% of my time marketing the next version or sectioning off this or that feature for commercial reasons.
But I take your point, I don't love subscription software either, but most of my current users are pretty comfortable with these so it's good for now while I'm working alone. And my hosting / API access fees are not _zero_ so there is a vague justification there too.