Surely the burden should be on the censors to establish clearly that something is in fact incitement to violence, rather than on external reporters to magically show that content which has been taken down is not incitement?
But the journalists seem to be doing a decent job of announcing and describing the data they have, and confirming it with multiple sources within Meta. They're engaged in a seemingly earnest and forthright effort to make the case. And to the degree that it's limited, it seems those limits are due to Meta itself.
Meta, on the other hand, excepting these whistleblowers, makes very little information available about their take-down actions both at the level of individual cases or at the level of their systematic responses to governments. The whistleblowers claim that Meta regularly took down posts without human review when requested by the Israelis. That's the exact opposite of the high burden of proof that you're asking for.
In terms of the implied proposition that israel is intentionally using the take down process to shield itself from criticism. I just dont think the evidence in the article supports that proposition. I would expect the stuff mentioned in the article to happen both in the case Israel is trying to get criticism taken down and in the case Israel is only interested in having "kill 'em all" type posts taken down. So i don't find the article very compelling.