The problem the article is addressing (I think) is that this reality does not seem to inform the popularly held beliefs and expectations of (inexperienced?) indie game developers, and the effects of that.
Reality doesn't inform a lot of creative endeavors. Writing some tech book (or really books more broadly) is almost certainly some way below minimum wage task from a financial perspective unless it makes your company notice you more, it brings you clients, etc.
You'd most likely be working on a small portion of a bigger game at a salaried job, without really being able to accumulate the needed skillset to jump into a larger game developed on your own terms.
Doing those mid-level games is arguably equivalent to taking a low-paid internship.
The hypothetical three year game made by an indie team expecting to make $300,000 is still going to be earning far below a normal software salary.
How long ago was that though? I started professionally in games in 2004 getting paid £16k/pa living in Dundee making J2ME games then moved over to another company making a UE3 title after six months going up to 19k/pa. I also had an offer for 21k/pa working for EA Chertsey but didn't fancy moving there!
I hope base salary has gone up by quite a bit in the UK since then!
2014. My salary has gone way up since then as I moved up the "ranks", but I know that company pays £25-28k to juniors straight out of uni nowadays.
A half decade before you, I also started at a big studio writing C++ (but with one degree) and was paid more twice that but in Canadian dollars (65k CAD). You're right that it pays lower than other programming jobs: Silicon Valley starting salaries were around 80k USD.
Am I understanding that right?
A reasonably competent developer could make at least twice that in salary, with far less risk. So that's not really profitable, once you remember opportunity cost (and especially if you risk-adjust it).
The ones that take a few days can be a weekend project. The longer projects you'd be hoping become huge, and very profitable.