I've been coaching for 17 years now and I can't remember when I started using autoregulation
(a-reg for short), for sure more than a decade ago. I have never used it in a sprinter's context though since I've never coached a sprinter. My own method, which I think is still a form of autoregulation, just not as much in the neurological sense, simply has an athlete look for signs of breakdown upon which the session is ended. If I recall correctly my first use of this was on one of my own hill interval sessions where my goal was 8 X 1:00 at max efforts. I thought it'd be a good idea to time the first one and also mark the start and ends and then run the rest of them by that distance aiming to complete it in 1:00. At #5 I felt a severe drop in strength and I ended the minute several meters short of the previous intervals. I ended the workout there. This is autoregulation. It's allowing your body to decide when to end the workout rather than force it to complete a predetermined distance or rep goal.
The concept tends to float around more in weight rooms. When you use a definitive load it becomes very simple to see the drop. Cycling is another sport where it's fairly simple to use because of wattage. In sprinting you tend to have to have a complex timing system that allows you to measure top speed for a predetermined period of time. It basically takes a snapshot of your fastest speed for the predetermined interval. For sprinters this is an interval lasting either less than 9 seconds, or one lasting 10-40 seconds. The physiological significance of those two times has to do with central nervous system stimulation or not. Maximum speed can only be held for 9 seconds in a well trained elite runner, for mere mortals it's closer to 4"-6". And on a side note, when you talk about the difference in training between elite women's and men's 100 meters, this becomes a huge deal. Men are hitting 9 seconds while women are hitting 10... that one second difference changes everything because of the CNS and it's 9 second rule. Anyway.
Today I decided to experiment with a 10 X 200 workout and allow my body to regulate the reps. I got to #5 before I felt the drop off. Although not measurable, I felt it for sure. My effort in hitting 30" was much higher and more importantly my form broke down. I ended the run there and moved into another system focus. Starting blocks. I still suck.
The next workout I will do is going to be 20 second intervals at max velocity. I'll be using a 6 meter drop off as my a-reg cap. I'll use 6 meters to simplify this and allow myself to measure it more accurately, but 6 meters equals roughly .6 seconds (3% of 20 seconds). Once I am unable to finish within that 6 meter zone I'll be done.
Why do this? It comes down to recovery. The last time I did the Sanya workout it took a toll on me. Traveling and all that is associated with it inhibited my recovery too, but that's part of it. Had I autoregulated that session I'm confident I'd have gotten in at least two more quality sessions since. Instead I missed them. Autoregulation is helpful in making sure you don't over stress your body and CNS and by tracking and logging the results you can eventually make sessions very predictable in terms of recovery.
One other way that I use a-reg with my athletes is on their long runs. It's not uncommon for me to put a long run as something like 10-18 miles. Pretty vague. But my notes on that sessions will say to let your body decide the distance. Don't force it. Let the miles come to you and don't chase the numbers. Once you begin to feel a loss in mechanics then head home. A loss in mechanics (poor knee lift, weak support in lower legs, plodding basically) is a sign that you have stressed your body adequately and anything more is a) probably not making you fitter and b) increasing your risk of injury. Again, I'm not positive that this concept truly fits into the original definition of autoregulation but I think it's close enough. It's still hitting at the heart of it.