I don't think it's a substitution. Lexicographically, there is a difference between an alternative (colour / color) and a synonym (hue). Now darn might look like a variant if you don't know the etymology, but ignorance is no basis for etymology. The substitution might abuse this fact somehow, but it isn't fooling nobody, is it? I hope the article will explain, but I expect they are blissfully unaware of how intricate soundchange is.

For example, dang seems to be a cognate of danger, a francophone form of damage, evident in quelle domage "what a pitty", very far away from damnit "may he burn in hell", but the details are hazy. So people err on the safe side of I know what you are doing with the minced oath. In reality, it's prudishness mixed with language racism, excuse my french.

> a pilot study of 100 people who spoke one of five languages that are only distantly related: Hebrew, Hindi, Hungarian, Korean or Russian

That's where I stopped reading. Because my statistics isn't up to the task of engaging their methods, and not because I were a militant defender of multiple language genesis. I just really need an excuse to spend my time better, elsewhere.