Well, even though my example is a relatively small change in behavior, it's the fact that it displays as a compulsion for such a long period of time, and is transcribed to a large number of different services, that makes me think that investigating the ways we design our technology can affect behavior. I've given a lot of thought to efficiency and layout before, but never how X UI would change Y thought and cause Z behavior which is then transposed to A + B + C related platforms, and D + E unrelated services, and real life on top of it.
I'd imagine that examples of a momentary, highly intense frustration due to UI/mechanics (i.e. dying in CoD, sudden loss of many days of work) are more likely to be turned into violent outbursts, while longer term frustrations (i.e. a shitty friends list UI, low rare item drop rate) turn into long lasting behavioral shifts. That's all guessing though, we need a lot more science to understand the impact well.
I think this article brings up something really interesting that I was actually prodding my friend about the other day regarding UI design. See, he was playing a game with a friends list, and he was telling me that he needed to delete friends. His list is far smaller than mine on this game, only around 40 people. I eventually dug down to his original experience with friend systems for video games - the original XBox. The XBox had an awful UI for sorting, displaying, and finding friends - you could only see 4 or 5 friends at a time, and it would never get a passing grade under today's UI standards. This was a system from nearly a decade ago, and there's a non-zero chance that his experience with the UI still affects his behavior a decade later, manifesting as a vague compulsion to keep his friends lists short.
So, how does this relate to the article? If a UI can train people into long-term compulsive behaviors, it's not unreasonable to research whether they can also nudge people's behavior in other directions on a shorter timescale.
No, he wasn't elected. He lost to von Hindenburg in the presidential elections, who later appointed him as chancellor. Meanwhile, the Reichstag switched to a largely Nazi composition due to success in elections, and passed a law (the Enabling Act) that gave Hitler the ability to pass laws without the Reichstag's approval. When von Hindenburg died, Hitler used the Enabling Act to merge von Hindenburg's former office with his own.
The only way you can claim Hitler was "elected" is by indirectly having his cronies get elected.