Have to be careful with these pronouncements about what is allegedly not possible, to be sure you don't claim something which actually happens in other situations... to be impossible.
Consider the idea the the rise in an ordinary waltz figure must not be before the end of the first step. Okay, sounds sensible, but when is the end of the step? The end of a foward step is when the other leg draws even with the newly placed foot.
But consider the man's action on step 5 of the spin turn. This is a heel lead, with a rise at the end of step 5. Only his left leg is at no point going to draw even with his right leg. Technically, the condition of the end of this step is not defined but it would have to occur by the time that the left leg gets as close to the right leg as it is ever going to. Since the feet are never going to close there, the rise on this step must occur while the feet are still apart.
Transpose the reality that it is possible to rise from a heel lead while the feet are still apart which we have just demonstrated in the spin turn, back to the first step of the natural turn, and that would be a rise well before the end of the step which in that case is not until the left foot draws even. That doesn't make it advisable to have a full rise that early, but it does make it entirely possible.
But to return to what is advisable, once the dancer's feet are strong enough, a significant percentage of top coaches will insist that the heel begin to rise from the floor ever so slightly before the other leg passes, such that the rise and resulting swing are fully underway by the conclusion of the first step, instead of delayed until what is really the early part of step 2.