On the article's point of "don't underdevelop," for launch I ended up trying to follow the SLC philosophy [2], a cousin of MVP. I managed to stick to the "S" with mixed success :-).
My own two cents for anyone out there thinking of trying your own thing: it's discouraging sometimes to read the flood of comments about survivorship bias, the need for luck or connections, the doom-y statistics about the percentage of companies that fail. Are these viewpoints based in fact? Yes, probably. Should you take the leap and try your own thing anyway? Also yes.
Reflect and decide ahead of time what your definition of success is for your venture. Money, satisfaction, adventure, idealism, hope: all are equally valid definitions of success, and just because someone may have a different definition does not invalidate your own. Work out your budget to see what you can realistically spend and for how long. Then go to work!
We are fortunate as software engineers that our field allows so much creative expression from just our brains and keyboards. Software can be almost a living thing, "pure thought-stuff, infinitely malleable" [3]. If your circumstances allow, it certainly deserves your attention, and the fact that sometimes it can produce economic value is a side effect, IMHO.
[1] - App for technical and engineering diagrams: https://app.vexlio.com/
[2] - Simple, Lovable, Complete: https://longform.asmartbear.com/slc/
[3] - Fred Brooks, The Mythical Man-Month, can't remember which essay.