It is one of definitely the most forgotten things from the past that has a lot of merits that today's instruction wrongly neglects.
Lead heel will always come off the ground unintentionally if the pelvis works in a slanted way (at 45 degrees up & back with the whole body and arms working together) and there is a proper sequencing (the lead side is totally passive during the backswing letting the lead knee joint bent inside). All great ballstrikers of the former years did it, from Vardon through Jones, Snead, Hogan, Middlecoff, Lema, Nicklaus, Trevino, Moe, etc. - the longer the swing arc the more visible it is - it vividly helps in achieving a full coil of the body as well as emphasize setting the vertical axis of downswing rotation in a correct time and place. Of course, we cannot forget about the correct weight transfer to the rear without which it would be difficult to expect the lead heel biokinetic dance, so to speak.
Somebody could ask - what about the theory of natural body limitations then ? Shouldn't the lead side limit the movement ? The answer is - the theory of limitations works best if it is not against physics. The firm rear side built sequentially from the ground up creates the natural human body limitations - while the lead side must be passive during the backswing so that it takes the lead later on when the orientation changes. Simple as that.
The lead heel should replace at least the same spot, or better said, a tad closer to the target because of the vertical axis of rotation linear shift to the front side of the body in a perfect world. If the heel replaces closer to the golfer's rear side most probably there is a spin-out in the pelvis area suggesting there is not enough big CoG shift. It also would depend on how much the lead foot was flared out in relation to the target line.
Lastly, one shouldn't forget that the pelvis motion is dependent on what happens below hip joints - and how the soundest biokinetically pelvis area motion looks like - please look at it through the prism of the lead heel motion. As we already know from the SPC concept descriptions and visualizations, the human pelvis motion during the golf swing always includes linear and rotational elements. We also know that the biokinetically soundest is to let the linear motion preceed the rotational one. Now, since noone on Earth could start the downswing properly with the lead heel off the ground - the heel must be replaced - and in order to do so, the linear part must be performed. After the heel is replanted all the golfer needs is the sequential rotation of the whole body system from the ground up. The biophysics is really simple as that. Too bad the modern instruction is so much blind at such wisdoms. Hopefully, it will change someday.
Look at the SPC pelvis area diagrams and confront them with what you have just seen:
As you can see on these diagrams, the rotational part always follows the linear part which is a biomechanical necessity for a bipedal. The key is to automate the cascade of events - both thanks to the theory of limitations - the linear part thanks to physical reaction and the rotational one thanks to anatomic reaction. The power of biokinetics.