There's a very simple answer to this. Their experts should provide a proof of concept implementation and then let everyone else pick it apart.
The criteria for success would be:
- It should not present a risk to that vast majority of internet users who are going about their lawful business
- It should not present a risk to those living under a repressive regime
- It should not present a risk to those, including but not limited t, children, living in abusive relationships seeking help
- It should not present a risk to whistleblowers seeking to draw attention to some illegal activity
- It should not present a risk to journalists working in hostile environments, including those working undercover
- It should not be open to abuse by unauthorised use or access by authorised users, including, but not limited to, Cheshire police intelligence (sic) analysts
If these experts can provide a robust practice demonstration of this they will have made their point, otherwise they, like the rest of us, should think of the children public at large.