People die. A key responsibility of being a parent is to make sure arrangements are made for you children should you die.

I suppose whether this risk is warranted involves some complex calculus taking, among it's many inputs, the quality of arrangements that the parent has made, the actual mathematical degree of risk, and personal factors including those about the ages, personalities, etc., of the children involved. And that's true of basically every activity involving risk, but I don't think the key parameters on the boundaries here are unambiguous even if the input values were clearly knowable and computation tractable.

In general, I don't think outside parties ever have the information to have much of an meaningful discussion on specific cases, and it mostly ends up being empty posturing.