At least part of the mismatch there, I think, is that being a well-respected researcher in most academic fields means excelling at an entirely different set of skills and activities than an undergraduate class typically covers, and having domain knowledge that is far more specialized than your students need to have. I mean, if I’m teaching a Shakespeare class, the vast majority of what I’m teaching falls into the general category of “stuff you should know to be well-informed about Shakespeare, but that nobody is doing actual research on, because it’s known already.” For example:
– How to scan a line of verse; how to recognize if it’s something other than perfectly regular iambic pentameter; how to think about those irregularities. And, more basically, how to tell the difference between prose and verse in the first place, and a little about how and why early modern playwrights use each.
– What an early modern theater looks like, how the performance conditions and conventions were different from today; what your part would look like if you were an actor and what kind of skills you would need to have to bring it to life on the stage. And lots and lots about the performance choices actors, then or now, might make in particular scenes from particular plays.
– Lots of foundational cultural / historical knowledge, like what sort of things you’d learn if you went to school in the Elizabethan era, and what we mean when we talk about “quartos” and “folios,” and who the Puritans were. Plus a little bit about what textual editors do, and how to tell, for example, if a stage direction is original or added by a later editor.
– And a TON about interpretative possibilities for the particular texts we’re reading: why, for example, you can read King Lear either as a defense of absolute monarchy or a radical critique of it, and where in the text you might look to support each view. (Mostly, I try to keep my own interpretative preferences under wraps. Very occasionally, I succeed.)
None of this is what the ground-breaking Shakespeare scholars are doing, and I don’t have a lot of the skills they do (heck, I can’t even read early modern handwriting very well). Some of it is informed by what they do – we know lots more about actors’ parts and cues now than I ever learned in college or even grad school – and more of it is knowledge / skills you need to have before you can become one of those scholars, but sometimes they aren’t that great at teaching this material at the undergraduate level because they assume everyone already knows these things, and forget their students don’t. But, at the same time, I wouldn’t be much use at my job if I didn’t have the skills I do teach down cold; if I couldn’t, for instance, look at a syntactically complex passage of verse and recognize at once what it’s saying and what rhetorical devices are being employed.
Whew, that was a dissertation? To get back to the actual thread topic, I expect there’s something similar going on with the sports-coaching example – you don’t have to have been a best-of-the-best player of a sport to coach it well, and that might even be a detriment, but I’d argue it’s not only about the teaching skills either: it’s also about having lots of less-flashy-but-still-fundamental domain knowledge about the sport, and that knowledge base probably has to be both broader and deeper than might be apparent from one’s playing ability alone.