A Proved Hypercorrection?

Here's a sentence from an article in The New York Times Magazine by Daphne Merkin. Note the use of "proved" as an adjective:

Writers in the business of trend-spotting have to come up with trends, after all, and Sheehy, the author of the best-selling "Passages," is a proved virtuoso at this genre.

People normally say "proven" in this situation. And I'm sure most people write "proven" as well. Is Daphne Merkin an exception? Maybe, but I have a strong feeling she originally wrote "proven" and it was changed by a copy editor who was misapplying the common prescriptivist rule that "proved" is preferable to "proven" when it's being used as a past participle (that would make it an example of hypercorrection). If Merkin had written this: "Sheehy had proved herself a virtuoso at this genre," then "proved" would have been "correct." But in what she actually wrote, it should have been "proven," as is clear from The New York Times Manual of Style and Usage: "as an adjective before a noun, proven is better: a proven remedy, proven oil reserves."

I was curious about how prevalent this usage was in the Times, so I did a couple of Nexis searches comparing "a proved" with "a proven" (that won't catch things like "we enlisted proven experts," but it does provide a basis for comparison). Turns out it's very rare. There are only two instances of "a proved" in the past five years, compared with more than 400 for "a proven."

Now, I'm not saying "proved" is certifiably wrong here, just that it's clunky and unnatural sounding. It looks like a case of robotic but faulty rule application. But it's not surprising that copy editors, who are trained to follow a set of arbitrary rules (even when they flout common sense), sometimes get confused and apply a nonexistent rule that turns something serviceable into something stilted.