Leadsled.
Folks,
And for damned good reasons, too, few GA aircraft are as reliable in recovery as a glider from UAs.
My understanding is that in Australia the 3000 ft limitation in the Regulations applies only to aerobatic manoeuvers where more than 60 degrees angle of bank are used. By this criterion, stalling practice is not considered an aerobatic manoeuvre.
Apart from some war birds such as the P51 Mustang, the manufacturer's POH do not offer recommendations as to minimum heights, so it is normally left to Regulators to apply limitations such as aerobatic flight. Individual flying schools are welcome to apply their own limitations on top of but not less than, State regulations.
Because someone stuffs up a spin recovery in a C150 and goes in, that does not mean every aircraft including an LSA must now claw their way to 3500 ft AGL to practice a stall recovery by 3000agl. Perhaps practice stalls in a steep climbing turn in Cessna 150's should be avoided due to the potential for a flat spin if poorly handled
It goes the other way of course. Obsolete aircraft like a DC3, in a full flap approach power on configuration, can drop a wing quite violently and it takes at least 1000 ft to recover to level flight even if you don't mishandle things. Been there - done that. I am sure even a qualified test pilot would want to have 5-7000 ft under his belt before stalling the DC3.
For Mary M. Thanks for your reply re glider operations and practice stalls. You made a most interesting point about tight climbing spirals in gliders at close to stalling speed. My point about an average loss of 50 to 100 ft in a stall recovery in a light single such as a Cessna 172 and flown by a competent pilot, was never meant to apply to spin recoveries - far from it..