Here's one example of a McArdle debunking:

http://inversesquare.wordpress.com/2010/07/24/why-friends-do...

Many more are lurking around the internet, should you feel inclined to dig them up.

I found that criticism less than persuasive. If a person says that $1000 in medical bills is to low a threshold to declare a bankruptcy to be due to medical expenses, then the fact that the actual threshold the study used was even lower ($1000 in medical bills or A or B or C) doesn't detract from the argument and merely makes the critic look bad.

All the other specific criticisms seem to be either equally pedantic, or matters where there could be a reasonable difference of opinion expect for his proper criticism of her for getting the response rate of the survey wrong which does significantly effect one of the points Megan raised in the article that was being debunked.

That particular example, though, is a strong one. McArdle wrote: "Warren and her co-authors defined anyone with $1000 worth of medical bills as having a medical bankruptcy". In fact, the paper used multiple criteria, one of which was $1000 in uncovered medical bills. This is not a pedantic difference, so far as I can see; there's a big difference between having $1000 in bills, some of which is covered, and $1000 in bills, none of which is covered.

McArdle didn't distinguish between covered and uncovered bills in her writing, and given the context (covered bills can't push someone into bankruptcy by definition) it was pretty clear she meant uncovered bills. So yes, the author is being pedantic for insisting that something that was clear from context be spelled out.