I don't agree in bold - you can only ever have visibility on a tiny percentage of cases, and nowhere near enough to draw meaningful conclusions. There's also obviously huge scope for bias to influence such a subjective method of assessment.
I think we're conflating two points here - my argument is that the whole 'London has fallen' narrative around crime and lawlessness in the city is grossly exaggerated, as evidenced by the data.
At a glance, I would agree that data also supports your suggestion that on average foreigners are more likely to commit crime, but I'd guess I think there's a little more nuance to that than you do.