OK, if we're being lazy, I'm going to comment on a part of the Wikipassage just based on my ambient knowledge of linguistic history. The line I am going to critique follows:
"A logical inference would surmise that the Proto-Indo-European origin for both is earlier than the linguistic split."
A couple problems with this: Both branches, Germanic and Celtic, went through significant linguistic changes developing from Proto-Indo-European. A *p- in that distant reconstructed language would have developed into an f- in Germanice (think of "father" versus words like "paternal"), and PIE *p disappeared altogether from insular Celtic(the Irish word for father is "aithir").
So "puck" looks like a "wanderwort," a word that has made its way into a number of languages in NW Europe, but probably later than say 500 BC, when many of these changes were taking place. Or, as the article cryptically suggests, the word was closely associated with place names, which tend not to change in the same way as other items of vocabulary.
Perhaps the word itself is puck-like, leading us down forest paths only to disappear without trace, leaving us bewitched and bewildered?