Popularity on a Google search of a software term is of course going to win the popularity contest, it's the Internet. Do the same search of economic journals and you'll probably find the bias toward "nudge" over "dark patterns" in an extreme way. Different ideas proliferate on different mediums.
I do strongly recommend checking out some of the Freakonomics podcasts[2] and blog posts that involve Thaler. Extremely interesting and useful across many disciplines!
[1] http://www.investopedia.com/articles/investing/102715/richar...
In this case, I'm pointing out that "dark patterns" already has a rich accumulation of psychological tricks on web transactions. Thaler can build on top of that shared knowledge. For example, people can go to darkpatterns.org and then click on "Browse Library" and learn many more variations of what Thaler is discussing. If a reader is only exposed to the phrase "bad nudge", he/she knows less instead of more.
Do people want to eventually converge on "bad nudges" instead of "dark patterns"? I don't care. But at the moment, "dark patterns" is the phrase that has attracted the bulk of discussion. (E.g. see previous cited google search links and HN itself[1])
[1]https://hn.algolia.com/?query=dark%20patterns&sort=byPopular...
For examples like those, I think the meme "dark pattern"[1] has more currency[2] and I wish the author chose to use it (or at least mention it) instead of expanding his "nudge" meme into "bad nudge".
An author will stick to their own terminology because of ego, or because of ignorance. I don't know which is applicable in R. Thaler's case so we can be charitable and assume it's ignorance.
[1] http://darkpatterns.org/
[2] ~263000 hits for "dark pattern" and ~930 for "bad nudges"
https://www.google.com/search?q="dark+pattern"
https://www.google.com/search?q="bad+nudges"