It comes down to this: if you ever found yourself in a situation where you needed to connect a USB device to a remote service, would you prefer to download that service's unsandboxed native code to your PC and execute it? Or execute some JS in the browser sandbox and grant it limited access to that one specific device?
I have never found myself in that situation. That sounds like a really silly idea.
For those that _do_ require [such use cases][1] though; they can now do so without needing to expose their system to an unsandboxed native app.
The web is a disaster and WebUSB is a prime piece of evidence supporting this.
While obviously we want to do as much as we can to discourage users from shooting themselves in the foot, there are limits. At some point, eventually you _do_ have to trust that the user knows what he's doing.
Giving users a choice on when to allow a page access to one specific USB device is not a "disaster".