As you're doubling down on this, I've no wish to continue here any more.
It's not big or clever to sterotype the Japanese.
A shame to leave such an entertaining interaction this way, but there we are.
What is that "this" I would have "doubling down"?
I saw a contradiction, and made it Captain Obvious with an analogy, because Captain Obvious Rules Them All in Making Things Clear.
This is a thread of 5-6 years ago, the case was a Snap after 3 years Service:
https://www.cyclechat.net/threads/shimano-105-crank-axle-failure.267532/page-2
The picture there:
https://www.cyclechat.net/attachments/485722/
... where Vapin' Joe Vaped the Thought I Thought after reading your nowadays, repeating story
Whatever minor "Benefits" we get with Hollowtech and the like, a few grams less weight and some perceived bollocks about stiffness I can't help thinking it is more trouble than it is worth compared to traditional square taper.
... and after putting figures on it (those online calculators), 20 grammes more (=reducing the weight reduction (=the "win" of the HollowTech2) towards square taper from 130 to 110 grammes) wins 29% better resistence to such torsion break, I'd frankly put how can anyone not agree?
20 grammes is within credit card weight range, and it buys a safety reserve of 29%.
The HollowTech2 design choices may be acceptable for competition racing in a case of replacing cranksets after every race, season, whatever within a couple years, because it can be claimed there that technology is a part of the competition, next to sport/fitness, the "best" being based on the two.
But outside that niche part of the bicycle users, I consider it quite Insane in the Brain to proclamate the design as a Standard for Everyone, let alone consequence of it failing is a recipe for injury.
The same can be said about their crank construction - instead of massive material, a construction of 2 U shaped cross sections leaving a hollow, together resulting in a part of that 130 grammes "gain" relative to square taper. It ALSO proved as failing, and in such numbers that it ended up forcing Shimano to do "something", despite far beyond warranty, as to avoid accumulated lawsuit costs that could end up far more costly than a recall, which has in itself a bogus factor, since inspection and seeing nothing guarantees nothing about tomorrow and any day beyond, proven by peoples stories, as I already mentioned here.
That implies a failure that propagates far enough without seeing anything externally. That, is the clear danger, THE recipe for disaster, since even people that bother checking get fooled into further using it.