I can't seem to figure out how such a bug would occur. Is it a coincidence that the sounds of the words are so similar, or is that what caused the erroneous results? I wouldn't imagine they would use pronunciation for the results but maybe for the query error detection (did you mean?).
This is a dangerous precedent. Google should (and probably will) appeal, and I really hope they win. Who knows, the "right to be forgotten" could turn into a "fight to be remembered" ala Trotsky.
I'm baffled.