In the case of this particular question, author Brad Phillips concludes that there's no right or wrong answer. I beg to differ.
As a writer, I've been on the deadline end of the wrong answer, and it stinks (to put it politely).
On Being a Good SourceMy answer to would-be sources is this: don't stretch for an interview unless you're really, r e a l l y going to stretch and actually become a good source.
Trust me, a prompt, professional response that you're not able to authoritatively address the specific aspects of the assigned topic will be appreciated.
Do you worry because you "hate to turn down a media opportunity?" Don't. Generally, a prompt, professional response of any kind will be enough to get you on the list of "people to call" for future assignments on any related topic.
As a source, the second-worst thing you can do is stretch a little. If you stretch enough to talk around the topic, but are unable to offer any real insight into the topic, you've wasted not only the writer's time but the readers' time as well.
The Worst SourceIn my opinion - as a writer and a reader - the worst thing a source can do has nothing to do with stretching. It has to do with bending.
Sources who bend, twist, or turn their responses, or who stray so far from the topic that their comments are useful only to themselves and/or their brand, are the worst... when you're prepared, that is.
Sometimes - rarely, but sometimes - a "bad" source can, like a lost sheep, be worth following. The trick is knowing when.
Don't Get Lost, Get LuckyFollowing a wayward source is worthwhile when you have enough time to appreciate the detour as part of the education process. Managing your work so that you can have "enough time" is a challenge every writer knows too well - and most know it depends on having a lot of discipline and even more luck.
To increase your chances (I'm talking to writers now, you source-types are free to leave) I've found it helps to identify potential sources as early as possible in your research phase. I've lucked out a few times by contacting those potential sources before I understood the topic well. In those cases, I suspect my totally honest approach improved my luck.
Contacting Sources Too Early Can WorkWhen contacting potential sources before I've had a chance to develop good questions, I say something like, "I'm researching X topic and would like 10-15 minutes of your time while I begin to understand how Y and Z are affecting X. Before you agree I should tell you this interview would be considered background only - in other words, no attributable quotes at this time. If you're willing to speak to me, please call..."
Obviously, when using this approach it helps to cast a wide net (or look for a lot of sheep). And it helps to have a long lead time - something almost always out of your control as a writer.
What is under your control as a writer, and what you must remember during the interview and writing process, is that you and your sources approach the interview from completely different perspectives. Put another way: as long as you understand that the question How far should you stretch in an interview? is one sources must routinely consider, it's probably unlikely that you'll be led too far astray.