Let’s set aside Roger specifically for the moment. But as a rough calculation lets call a mile 4 laps of a 400m oval. And to run four laps in 4 minutes you need to average 60 seconds/lap.
As you say a 400m is about as far as it is possible to sprint. I think physiologically it is shorter but, back of the envelope, 400m is close enough.
So if 400m is the maximum sprint distance every distance greater requires some measuring of effort? Even if to mere mortals people running a mile that fast look like they are sprinting, they are actually running sub maximally. If they weren’t they would slow dramatically before the end of the four laps because they have misjudged their effort.
A four minute miler can probably run one maximal 400m in the range of high 40s to low 50 seconds. So that would be their maximal intensity. Anyone trying to do that 10x would, like you say, be in a world of pain.
10 x 400m isn’t a maximal workout, it is what is needed to meet the demands of a four minute mile. A miler that fast could do 12, 15 or more if it was a ‘how many 400s can you do in 60 seconds’ competition. But it would have been a measured effort done to meet the demands of what he was trying to achieve, without injuring himself or getting sick etc.
Intervals are just a means to an end and a good coach know show to manipulate the length, intensity, rest between and any other variables to achieve the desired training effect on a runner.
For the record I can’t run one 60 second 400m :( I can run 10 at 75 seconds a pop because that’s what someone at my level of fitness needs to run to reach my goal 10k time.
Not to the marathon yet, but in the late 50s and 60s a guy from NZ, Arthur Lydiard, took inspiration from Bannisters training but proceeded it with a period of long slower running, a hill sprinting segment (HIIT like maybe), then the intervals, and a taper. His athletes won about 6 medals over two olympics including Peter Snell who won two golds and has a mile best of 3:54. Some consider this ‘linear periodisation’ through phases to reach a peak the start of modern training (that’s a discussion for another day).
Your question about the difference between what coaches do and science is an interesting one. A recent well regarded running book ‘The Science of Running’ spends the first half diving in to the science of biomechanics, metabolisation of oxygen, lactate clearance and all that good stuff. The second half focuses on the training of athletes to achieve success in specific races. The stereotype of ‘Coach’ as a barrel chested, loud mouthed, ‘run till you drop dead’ guy isn’t the reality these days. Most college and professional coaches have extensive exercise exercise physiology and/or biomechanics backgrounds etc.
The discovery of physiological thresholds like aerobic threshold, lactate threshold, VO2max etc has been invaluable to training but they are proxies for performance. And at the highest levels differences in performance can’t be explained away by slight differences in these markers.
Aren’t you pleased you’ve never sat next to me at a dinner party?
Lastly, HIIT and the marathon. HIIT is used in marathon training, but often not close to the actual event. One modern training philosophy is that to run a certain event you need a spectrum of qualities from pure speed at one end, the specific combination of speed and endurance you need in the middle, and pure endurance at the other end. When you start training you start at both ends and work towards the middle as you want to slowly close in on the specific blend of those ends of the spectrum. This would look like very fast high intensity hill sprints with long rest (10 x 10 seconds with 3 mins rest). The long rest is important as you need to be fully rested to produce max force. On the other end of the spectrum is easy jogging for maybe 2 - 2.5hrs. Slowly you drop the intensity but extend the distance of the fast stuff and speed up but drop the length of the slower stuff until you have the specific blend you need to perform.
This is applied slightly different in the marathon , but close enough. The problem with doing fast HIIT running close to a marathon is that a marathon is an extension of speed and endurance, but is also limited by the amount of glycogen that can be stored in muscle. Running aerobically burns the glycogen more slowly and if this speed is exceeded the rate of glycogen usage goes up. A big part of marathon training is running at a level where your body learns to use glycogen efficiently. If you do fast training you signal to your body that conservation of the glycogen isn’t important, and it can be sacrificed to run more quickly.
That said, for an untrained or slightly trained person who wants to run a fall marathon I’d 100% include intervals (maybe without the HI, we won’t agree on that) as the benefit they would get from them outweigh optimising fuel usage on the day.