I was really lucky to have an available, hilly, competitive group ride. My first time out with them, I realized that the idea was to TT the ride, i.e. ride the course in the least possible time, duh. How to do that wasn't immediately obvious to me but I figured it out on my first ride with that group. The power necessary to maintain a speed on the flat is proportional to the cube of the speed. So what one does on a hilly course is to ride the flats in Z2 and give it all you've got on the climbs. If you get dropped on the climb, hold some power going over the top and you'll get them back. Thus TTing a course is exactly what one needs to do to get results. A caveat: our routes were chosen so that we climbed from 50'-70' per mile and our routes were in the 60-70 mile range, so no more than about 4 hours. Longer or more climbing than that and the usual cyclist can't hold Z4 on all the hills - one gets too tired. We also had a contested sprint at the end of every ride.
It turned out that this was, IME, all that was needed to ride centuries, doubles, and even 400k was those 4 hour rides at one's limit. "If you can dismount normally and walk comfortably at the end, you could have gone harder." We also did non-competitive long rides in the mountains to figure out what long distance efforts felt like and to get our hydration and fueling down.
We had a rather ordinary distribution of effort by zone, about what one would encounter on a well thought-out training program, but we didn't have to fool around with the usual planning of zone-based training, it just happened. During the week, we'd just ride Z2, get some more miles in. It seemed to me that averaging about 100 miles/week was enough. Some people would average much more than that, but it didn't seem to increase their performance on the long rides.
__________________
Results matter