For example, how much would I personally for a fix to a graphical glitch fixed in the Dolphin emulator? Think how hard that is to quantify! Is it worth $20? It it worth $0.00001 over thousands of users? Is it not worth anything because nobody even noticed it?
If there was a dollar amount attached to it, would it ever get fixed at all? Maybe it's value is too low to bother now that we've quantified it that way; new features are worth so much more.
Even if you sell something for what's it worth to the end user, it might not be worth doing. Most of my open source projects fall in that category: It's worth maybe $20 to an end user, and there might be handful of users, and that's definitely less than my hours are worth to make it. But that's why it's not a business.
Sure, but that article mentions business (you do too in your comment) that expect something from the OSS project. The alternative for that business is certainly much more expensive than paying for that fix, thus it does seems to be worth that much to them.
That tiny fix could take minutes to a person that knows the code base, and weeks to one that doesn't. Its financial impact could be in millions.
I agree with your general thrust but on a tangent, this isn't true and isn't the right way to think about software contributions, even 'trivial' ones.
A trivial change seems low value to you because you wouldn't have to spend much time at all doing the same thing yourself.
To see its market value, you have to view it from the perspective of the average person - how long would it take them to make that change? Answer: a very, very long time. They'd have to learn the fundamentals of programming first.
The value of something isn't how long it takes to do it in isolation. It also factors in how long it took to get to the position where you could do it at all.