I'm not sure which method they used to black out the text in this case, but the classic stupid move is to set the background to black behind black text. If "you printed out the [document], or brought the file up on your computer to read", you couldn't read the text. But nothing prevents you from selecting the text, causing it to reverse color and become visible; and to copy it. That may be what happened in this case, if the text was still visible to a search.

As a general (universal) rule, digital documents are chock full of metadata that can give away the game if you have something to hide. Like the author's name; wasn't that one of the things that got Manafort in trouble with PDFs?

Another classic trap is geolocation in images. I had a buddy a few years ago send me digital images of his cannabis crop, I checked, and sure enough the latitude and longitude of his plantation was embedded in the images. I was on the phone instantly.

I'd say by now most smart journalists check public documents for metadata and stupid redaction tricks. I know I do, and I'm not a journalist, and my intelligence is often disputed. But the truth is in there.