+1
The argument that most resonates with me for double threshold training is the consistency it enables.
Everything about training involves tradeoffs. On a per-mile basis, lower intensity provides less stimulus that will produce adaptations for racing fast, but you can do more of it. Higher intensity provides more stimulus, but you can't do as much.
The double threshold training model seems to check the following boxes:
-It is fast enough to provide decent stimulus on a per-mile basis
-It is slow enough that athletes can do a lot of it
-It is easy to supplement with training on either side (like a weekly long run, even if it is only like 80 minutes, and a session of 200 meter hill reps) to fill in the gaps
-When athletes stick to the assigned intensity, it is highly repeatable, necessitating less down-time from real training in the long term. When combined with a long run and some limited shorter work, it is also going to get athletes close to top race shape, so huge blocks of higher-intensity training are unnecessary
All of this means that over the long term, athletes can pile up a huge amount of stimulus and adaptation (maybe even the most?) that leads to them being extremely fit and racing fast.