There are two ways to quantify risk, though. One way is about how serious the risk is. A papercut = small problem. A stranger could potentially kill or traumatise a child, which is obviously a very big problem.
The other is about how likely something is to happen. A papercut is very common - most people get several a year. Being murdered or abducted by a stranger is extremely unlikely.
I think the poster saying that a risk is tiny was referring to probability and not severity.
We aren't very good at judging risk as humans, though. We put way more emphasis on things which are on the extreme end of the severity scale and this seems to diminish the importance of the probability factor. In general, I don't think humans are well wired to judge probability - we seem to use our own experience, plus learned experience from others, in order to do this, and that doesn't map very accurately onto the big picture. The other problem with this is that because we consume news media, we get a really skewed picture of what the probability is for various things because we didn't really evolve to understand the world in that way.
If you think back before mass media existed, you'd hear about terrible things which happened in your own immediate circle, your physically local area and then maybe second or third hand experiences from others - that would be about it. Our brains still react to news reports that we hear as though this is the scale we're hearing about it from. So when we hear about two high profile child murders in a year, it basically makes us believe on some level that child murder is much more common than it really is because it's as though we are taking in an account from somebody who is physically or emotionally close to us, rather than being a representation of something which happened in another city. If you think about something which is less commonly in the news, but affects a couple of your social circle each year - maybe losing a parent, or being diagnosed with a serious illness. We can come to see child murder as being as likely as those things, and that is terrifying. Of course you'd want to reduce that risk.
In reality though, stranger abduction is so rare that it basically makes no sense to worry about it - it's about the same as the risk of dying from a general anaesthetic, or the risk of death from skiing or horseriding. It is a real thing which happens, but most of the time we don't think about it and don't consider it a likely outcome - we would only avoid that thing if for example the person was especially vulnerable (like a very young baby, or letting someone go out skiing with no training and no protective gear).
It makes a lot more logical sense to worry about things like childhood obesity leading to future health problems, difficulty socialising leading to isolation and depression, addictive behaviours possibly leading to drug/alcohol abuse. But because those things don't register as being as urgent as children being killed/abducted by a stranger, we consider strangers the bigger risk.