Those are all examples of stories complaining specifically about facial-recognition working worse for some races. The quantity, and prominence of the institutions spouting them, is the point.
Which data backs up your point?
My evidence isn't perfect - for that, I'd have to collect an unbiased sample of anti-surveillance stories, and categorize them according to which object to surveillance universally, and which mostly object to only discriminatory surveillance. But the impression I got through casual web browsing is that the latter significantly outnumber the former.
It's actually a very solid argument against the charge of having offered a Strawman.