There are always exceptions to any rule but the FAA appears to agree with me that this is not a good idea. I found several articles on the subject including this one.
The Impossible Turn (DOCUMENT FAA-P-8740-44 AFO-800-1283)
No doubt about it, in the vast majority of cases, this idea does not work out. Primarily because with most light singles, you don't have a prayer if you are not at least at TPA, and secondly because maneuvering at MCA is not practiced much after the PP is earned.
Most of these poor souls don't come to grief because they make the turn without putting a wing into a tree, yet still overshoot or drift downwind... invariably they stall in the turn and spin in. A power plane will need a lot more altitude than a glider to do this, but the basics are the same: 1) nose below the horizon 2) coordinated turn, as steep as the airspeed allows, into the wind 3) slip/flaps/spoilers as needed.
That being said, my personal rule when flying with an engine is that if I am below TPA, I might turn a little this way or that, but going back to the runway is not an option. I've tried "the impossible turn" at a safe altitude, and while it's surprising what the average Cessna can do, it's not worth it close to the ground. It's a lot easier to do a good job of turning engine-out with minimum alt. loss, and without stalling, when practicing than it is to do it when the fan quits right after takeoff!! :o
Besides, if there is traffic departing behind you, even if you pull it off- what then?

There's the rub, even if you are really sharp and manage to turn around and get lined up. At 47N, there is the option of the grass... but there are also good spots just beyond the boundaries that don't involve hairy low-altitude turns or creating a hazard to others. So I wouldn't even consider it. If I have 1000 feet or so (in a 172 or similar), I think I could turn to join the downwind and at least make a 180 to land halfway down the runway, into the wind. But again, there's traffic to consider. They could yield, but it would be risky.
So I agree that in most cases, it's best to head for a place you can see, with as little maneuvering as possible.
Those engine-failure scenarios usually work out much better.
