You know, before actually looking at the research I'd have agreed with this post. But, since having read through a few articles (during a couple extensive, and very enlightening debates about this subject) I'd have to say that this is not in actuality what occurs.
Here is the thing, like Zatsiorsky mentions, there are basically 3 ways to recruit maximal MU's/muscle fibers.
The first is to lift a maximal load (all fibers must be called on in order to produce sufficient force), move a submaximal load at a maximal velocity (according to the research I've read, ballistic contractions lower force thresholds, thus leading to Mu's/fibers being activated which would have have been at submaximal velocities), and move a submaximal load to momentary muscular failure.
(fatigue also lowers force thresholds, thus once again causing the body to activate larger fibers which would not have otherwise been called upon)
The most interesting thing that I noticed in regards to the fatigue study was that no fibers that were ever called upon ever dropped out. In other words even during those reps at the end of the set where your speed decreases, your largest fibers are still firing.
It's just that due to fatigue (build up of lactic acid, lack of available energy, etc...) those fibers are capable of producing less force, thus the speed slows down.
Now, am I suggesting not to use fast lifting (concentric) speeds? No, absolutely not. But, the idea of avoiding failure (or slower rep speeds) doesn't make all that much sense to me personally. And like someone said above, perhaps CW can clarify his reasoning for this in a future article or over in his "Maximal Recruitment...." thread.
Another thing that Zatsiorsky mentions is that in order for a muscle fiber/MU to hypertrophy, it must not only be recruited, but also fatigued. So, why on earth would one want to avoid fatigue? So you can do more sets/reps to reach the same level of fatigue? Does that really make sense? Is that really a more efficient use of time/energy?
Seriously, think about it, you have two options (and while these are hypothetical, there is some real world anecdotal evidence for both);
A) you can perform one all out set to momentary muscular failure at which time you have fatigued all of the muscle fibers/MU's capable of producing enough force to overcome the resistance.
B) you can stop your sets short of momentary muscular failure (using speed, ROM, etc... as a gauge) and then perform multiple further sets until your muscle fibers/MU's are fatigued.
So, I can do one set or five to achieve the same basic objective. Now, I'm no rocket scientist, but it seems to me like doing one set would be more efficient.
Now, before you all jump on my back and chastise me for being a HIT Jedi, I most certainly am not. But, I can see why they went the direction they did and the basic theory they were operating from. After all, why do we perform multiple sets? Why does volume have any bearing on hypertrophy at all? The answer, fatigue and muscle damage.
If that can be achieved using one set (or one extended set, say drop set, rest-pause, etc...) why prolong that fatigue and muscle damage just so you can do more sets?
Maybe I'm way off. Maybe I've completely missed CW's point and he can set me straight. Maybe one of the other forum members can shed some light on this subject for me. But at this point, while I like the concept of lifting as fast as possible, I don't understand nor agree with stopping the sets when speed slows down.