Don't agree with article. Parameters that can be tuned are not mathematical abominations.
In a mathematical model there is often a single value for a parameter that can be shown to be optimal. In the complexity of the real world, tuning is often necessary and an optimal number can usually not be proven.
Second parent, I don't agree with the article either. Real world engineering is not pure theory. It is a subtle difference. A perfectly valid reason for this type of problems is "it works well in practice", or "it is good enough".
The quality of the article can be summed up by one of its sentences: The consequences are dare.
You could try to reverse engineer that sentence to figure out if it came from "truth or consequences" or "truth or dare", or you could just flag it and move on to something well-written.
In a mathematical model there is often a single value for a parameter that can be shown to be optimal. In the complexity of the real world, tuning is often necessary and an optimal number can usually not be proven.