For Northern Star : I don't use microcycles in the classic way. I prepare training plans lasting 2 weeks, and update the next looking at the REAL effects of the present period. Using this system, I don't go to calculate the percentage of the speed basing everything on physiological tests, but I use a mathematic approach. So, if I see the athlete able running 7:35 at the moment, and one month later I want him running 7:30, I try to develop the specific speed endurance, using various distances between 90% and 110% of that speed (in this case, 15" every 100m, with my system of calculation 110% in 13"5, 105% is 14"25, 95% is 15"75 and 90% is 16"5), of course connected with the distances.
For example, at 110% I can use 200m (27") and 300m (40"5), at 105% 400m (60"), 500m (1'15"), 600m (1'30") and 800m (2'), at 95% 1000m (2'37"5), 1200m (3'09"), 1600m (4'12"), 2000m (5'15"), at 90% 3000m (8'15") and continuous run.
I gave an example of percentage, but in reality there is a easy decrease of speed, without jumps of 5% (for example, 1000m in 2'32" is about 98-99%, and we use frequently).
At the end, is the personal sensibility of coach and athlete which decide the real time.
Preparing 3000m and 5000m we use the same base as volume and intensity. The difference is only during the last specific period of 4 weeks, when we go for 3 (maximum) sessions of specific speed, and, if we look at 3000m,we use some less volume (for example, 6000m instead 8-10000m of tests on track) opening recovery and running faster.
About the long run in winter, Caleb never runs longer than 1:30:00, and also this is not frequent. However, he runs several times 1:15:00 / 1:20:00, using good speed (faster than 3:30 / km). Particularly, he uses "medium fast continuous runs" between 12 and 15 km at a pace of 3' / km or faster, runs some cross country of 12 km, and runs one 10000m on track in Kenya for controlling his endurance (this year was a "solo" in 28'28" in Nairobi on 10 May, with the second 35" behind).
With athletes as Caleb I don't need to have very long run at easy pace. He is very tall (188cm), and with long slow run becomes heavy in his action. Very long run with tall athletes, not able to use high frequencies in their strides, is Always a mistake, because provokes big problems under biomechanical point of view.
The goal is to create the capacity to run long and fast, may be once every month. Shaheen, for example, one week after winning Golden Gala in Rome (July 2005) in 7'56",
ran in Iten 37 km in 2:01:00, 5' faster than the Marathon runners of Kenya Team for WCh. But the long run of Shaheen, as the long run of Caleb, always was fast, under 3'20" > 3'10" per km (many times under 3').
This means nothing to do with the classic "Lydiard plan", if somebody thinks the situation is similar...