I generally prefer a combination: crab for final then somewhere around flare height change to wing down.
In some particularly gusty/strong winds the a/c I fly (an Islander
) is incapable of being rolled fast enough to vary the sideslip to stay over a particular point, and is also too slug-like to be yawed fast enough for a standard crabbed landing. When conditions are like this I leave the yaw until an instant before touchdown then used maximum rudder authority + asymmetric power reduction. I still tend to want to lower the wing though...
Chuck, I think they're referring to IF the student should stall while in a sideslipped condition due to not maintaining the correct speed etc.
Not sure why increasing the speed by a knot or two as was suggested would be beneficial unless they're compensating for pitot-static errors caused by the slideslip...