Some further thought:

Security threats in places less safe than Europe

I initially haven't thought about security threats of altruists in less safe places than my own. From my times at Amnesty International I know that activists and journalists in for example basically all of Latin America, Russia, Turkey, Saudi Arabia and China have to face severe risks to their personal safety. It might furthermore be less costly to hire security in lower-income places. I have no further insight here, but it might be fairly easy to find key people in such areas that do highly valuable work and only due to lack of funding can't afford the safety that would be efficient here.

As a concrete program related to the safety of altruists, I just stumbled over the German Elisabeth-Selbert-Initiative (they got a shoutout from the new German government). It is intended to support, protect and offer relocations to Germany to threatened human rights activists. (link)

Their programs:

PL1: Four to six-month relocation grants for human rights defenders at risk at host organisations in Germany
PL 2: Four to six-month relocation grants for human rights defenders at risk at safe places within their home countries or regions
PL 3: On-site assistance to protect human rights defenders; programme

I wonder how robustly valuable people here think this is.

My feeling is that this is worth looking into more.
It seems to me that there has not been a lot of public written analysis of risks associated with poor operational security in each cause area and operational security in the EA movement itself.

One possible explanation for this would be that some people have done this analysis and decided to keep the results confidential.

Still, I would also be very interested to hear if people know of such analysis being done.

With regards to death threats being realized: From my experience in Colombia, while most death threats do not get realized, they still act as a strong deterrent, often by targetting family members instead of the main target. 

Hiring a security detail is not the only possible counter-measure though: The main focus of the NGO I was working with was to increase the political costs associated with this kind of violence, thereby decreasing it's cost-effectiveness.

 

Polling the community in the forum

While thinking about what kind of feedback I would find useful for this question (after a couple of people having upvoted and no comments so far), I would've found a cheap poll with options like "Definitely not worth looking into this further" or "Probably uninteresting, but an interested EA might look into this more" very useful here. I wonder if this was discussed before, seems like an easy to implement and useful feature for quick and dirty feedback. Maybe downsides could be giving a wrong impression of EA consensus due to selection effects (e.g. most informed EAs being less active readers of the forum), or less in-depth discussion because people that would otherwise have shared their thoughts now only participate in the poll?