Technique versus parlor tricks
There are two concerns I have with this technique. First, it seems to rely less on solid technique than on social factors. Second, the set-up is transparently counterfeit.
In judo, I get used as uke occasionally. I take kumikata grips or an unusual grip we're working on, I stand with good posture or one of the commonly-seen-in-sparring poor postures, I move with good footwork or take a specific misstep, et cetera. The instructor uses me as a dummy if we're doing technique straight from the gokyo (e.g. osotogari). If we're working counters, set-ups or tactics, then a specific technique follows from one of my specific mistakes. I'm compliant, but the technique we're learning is rooted in usable, tangible, functional application.
There's also no hand-waving about how to translate the technique into practice. Either we're learning a throw that we know works because millions of judoka throw their trained, resisting, motivated opponents with it in competition, or we're learning a tactic or counter that we can test ourselves in randori later that night. The set-up has very close to a 1-to-1 relationship to application, so we know (or can determine) its value.
When I trained in light-contact karate, I'd also get used as uke for demonstrations plenty. Sometimes it would be similar to my description above: I'd punch, instructor would show an evasion and counter. Straightforward technique, straightforward set-up.
But much more frequently, as is common in non-sparring or light-contact arts, the set-up would be quite strange. It would start with "grab my wrist", and there'd be an unspoken subtext of "and don't react like a normal person when I start applying this convoluted, multi-step technique". Or, "grab my lapels...no, not like that! With your arms outstretched like a zombie, and now just stand there." Set-ups like that aren't just compliant. They're contrived, and often elaborate, awkward, and unlikely to boot. They rely on situations that don't actually occur (or occur like that) in fighting.
Also from my experience in light-contact karate, as well as eyebrow-raising demos from aikido instructors, I'm wary of techniques that seem to have over-the-top results. Sometimes, the technique is valid and just unusual. Other times, the technique simply does not have a solid basis. Instead, its production of over-the-top ukemi relies on cooperation from a compliant demonstration partner.
The technique is called "jump for sensei"
Okamoto is working with incredibly dedicated, long-term students of his. They are demonstrating for a crowd. They are being taped. Both uke and tori have invested years, if not decades, in making these techniques look effective on a compliant partner. In the past, it is almost certain that if Sensei attempted a technique and uke did not flip over and tap, Sensei would apply A) some other technique with a greater degree of force, and B) group social shaming of that individual for not receiving the technique compliantly.
Gillian Russell's Epistemic Viciousness (PDF) makes this point eloquently. In part, she says:
Not everyone treats their martial art like a religion, but another, more inevitable
problem for martial arts epistemology is that those who already have beliefs in
the area tend to have a lot invested in those beliefs. The people whose testimony
we are most likely to believe have inevitably put years of effort into perfecting
Everyone involved in Okamoto's demo is subject to an extraordinary degree of cognitive dissonance of the form "whatever Sensei does, I'd better take ukemi, and fast."
More than anything else, Okamoto has set up a contrived parlor trick. He twitches his hands and the students jump for him. This is meant to showcase some broader principle or efficacy of technique, but largely it demonstrates the power of social suggestion and long-term setting of expectations.
For an example of how this fares against actual martial techniques, please see this challenge match (YouTube) of a "kiai master" against someone trained in punching and kicking.