I suspect the point is that close to the ground one tends to use ground cues to a greater extend, rather than pure attitude cues (which are less reliable because of terrain changes) and instrument changes (because lets face it, we should be flying with our heads
out
in the circuit).
I found myself getting mildly thrown a little while ago in a slow (45kn) aircraft as I turned downwind (wind was about 15kn down the strip). Once I disciplined myself to maintain attitude, and not do what "seemed" to look right from the ground cues I found my airspeed remained fixed, just like Hudson. I could see how easily I could have stalled, but it was me, not the aeroplane.
It would be an interesting trial to do (with a safety pilot with their head inside, plus an aircraft known to lose very little height at the stall), ask various pilots to turn downwind at around 600 - 800 ft with a reasonable wind down a friendly runway following ground cues alone and see what happens.
G
[This message has been edited by Genghis the Engineer (edited 03 February 2000).]