The humans in question are dangerous, they're not civilized (so they're wild), and they're animals, so yes, they are.
> In humans dangerous wild behaviour is a (hopefully, and ususally) more or less temporary and fixable aberration
This isn't substantiated by research. Quite the opposite actually, psychologists have found that no intervention is effective in modifying violent antisocial behaviour in young males (and it's almost entirely young males). The only solution seems to be incarceration until around the age of 30, at which point most people simply age out of these behaviour patterns.
Also, there are many dangerous animals that can be rehabilitated, pet dogs for instance. It doesn't stop us from mandating they be put down.
> as a means of maintaining a human(e) society it seems rather ineffective.
I'm asserting it's effective at maintaining peace and order. I'm also asserting that the fact of peace and order is more important than the treatment of those who violate it.
And I could assert quite a lot about your assertions, the mindset behind them, and what kind of society that leads to.
But I can't be arsed to: No matter how many deluded authoritarians you manage to bring to their senses, or just reduce to a heap of smoking online cinders, more fresh ones just keep popping up. And I got bored with Usenet flamewars decades ago.
Because they aren't dangerous wild animals, while dangerous wild animals are. In humans dangerous wild behaviour is a (hopefully, and ususally) more or less temporary and fixable aberration, while in many kinds of animals it's the immutable default.
That difference shouldn't be too difficult to grasp, or is it?
> Is immediate and severe escalation of violence in response to aggression ineffective?
Not at fixing the immediate danger, but yes, as a means of maintaining a human(e) society it seems rather ineffective.