It's pretty clear that the left ends of the horizontal lines are vertically aligned but the right ends are not. This is confirmed with a ruler. Measuring horizontally from white to white, on my screen the top horizontal line is 50mm and the bottom one is 52mm. If I measure between some estimate of the visual centres of the junctions, I get 48 and 51mm.

I find the moral example no more illusory. It would take a large amount of research to draw the connection expressed by the "therefore". None is given. One can easily come up with hypothetical causes of the alleged statistic (alleged, because one would first have to ask "Is this true?", as I did with the lines). There are references in the Wikipedia article from which this is quoted, and one might start there if one wished to form a real opinion, rather than blow with the winds of other people's chatter. But the quote on its own is worthless to that end.

I'm not sure that the Africa one is a good example - it depends on your definition of the term "biased". if there are an equal number of criminals in Africa and Europe and the ICC is only investigating Africa, then that suggests that the ICC has some sort of bias towards only investigating in Africa. Perhaps the people in charge of the ICC only think that African people can be war criminals - then they're certainly biased against Africa in one sense, even though their actions are helping Africa. On the other hand, if their actions are helping Africa when they could be helping Europe, they're biased against Europe in the sense you meant.

A similar example is with police investigation in black communities in the USA. Suppose for the sake of the example that there are an equal number of criminals in black communities and in white communities, but the police investigate the black communities at a higher rate. If they do that because they think that black people are more likely to be criminals (even though this is not true), then they're definitely biased in the former sense against black communities. And in fact, they might catch criminals more quickly there (good) but they might also shoot innocent people (bad), meaning that they could also be biased in the latter sense.

I suspect that when many people say "biased" they mean the former sense, where it's about the perception, not about the outcomes.

However, I do agree with your general point. I hadn't really considered before how moral statements could be illusions in such a similar way to optical illusions.