However, in both cases the official made a quantitative argument, not a moral one. Some of your examples also aren't convincing: there is a lot of evidence that after controlling for socioeconomic status and other similar factors, black and white people are the same, and that rape is traumatic (rates of depression, PTSD, etc. following it). And personal feelings actually are admissible evidence, if the questions you're asking are designed correctly. The 'right to freedom' and valuing people with physical or mental disabilities are moral issues, but they are also fundamentally different from the claims the official was reported to be making.

Something like 'longterm damage' is something that can be quantified and measured (correlation, at least). There are a variety of psychological examinations (such as those for trauma, underlying biases, etc.) that you could use to see if early (or earlier) self reported access to pornography in fact correlates to different mental/emotional states later in life. At that point you have quantified changes and can point to them and argue societal values.

The moral argument would be "Children should not be exposed to pornography simply because it is not suitable for people under (age)", not positing things about potential, unquantified 'longterm damage'. We can argue all day about the morals and societal expectations, and it's perfectly reasonable to do so with regards to pornography, but if someone makes a quantifiable argument, they should provide evidence to back it up, or a compelling reason as to why evidence isn't available and why acting without it is still important. (Such as how long it might take to get the evidence, or why it isn't possible to, or why there is a dire and immediate need to act before evidence can be gathered.)