I think it starts becoming unclear in the $200k-$500k a year of donations range

I have slightly mixed views on this depending on the method. I in general think that competing in very typical fundraising methods, EA areas will typically underperform. Something like door-to-door or online ads - I expect EA areas (picked based on effectiveness vs appeal to donors) would pretty constantly lose to other NGOs who do a good job. On the other hand, I do think EA can have a pretty strong competitive advantage working on things that are closer to common goods that a single charity might not do itself but makes sense if you care about multiple charities/cause areas. E.g. when I go to a philanthropy conference it tends to go better than a classic charity as I am pretty comfortable talking about 50 possible charities with other donors instead of 1 which NGOs themselves might focus on. That just gives you more surface area to connect. Similar things like charity evaluators or setting up philanthropic funds I think EA might have an advantage on relative to classic NGOs.

If anything I think charities might be a bit too advantaged by going through AIM. The success rate is about ⅓ in terms of CE charities becoming field leading - this is compared to something like 1/10 of charities founded by other founders in the movement/via other incubators. I think part of this is fairly earned (e.g. strong cofounders and ideas selected via more rigorous process) but part is reputational - a bunch of funders I know are categorically more excited about AIM charities than identical charities that have not gone through the program. I think the same goes for hiring high talent staff. Why that happens concretely, I think cofounder pairing and idea pairing create a huge value add and the training, mentorship and community help new charities avoid some predictable mistakes.

I like meta and I like HNW broadly as areas partly due to there already being a solid number of country-level effective giving organizations. More specific ideas within other comment.

I am very optimistic about promising things being done in this area. I think being able to talk genuinely about multiple cause areas and being less directly pitchy on one worldview/NGO is a huge advantage a lot of EAs have. We are running a round on improving philanthropy and I would guess ideas we recommend there will be UHNW leaning. I wrote about a couple to do with dessert events and dragons den. Other things that are high on my list right now include grantmaker and philanthropic advisor training (a bit like our grantmaking training program), funder networks (like FAF but for other cause areas), more cross-cutting quizzes/tools (like GWWC's how rich are you tool). We are putting pretty active research into this over the next 6 months so I think I will have a better answer then, but on net, I think it's quite feasible there will be like 6 ideas we would be excited to see a full charity on - maybe 3 on HNWs.

Definitely an entrepreneur with no research skills. Given that most folks pick an idea our research team already looked heavily into, research skills are a bonus but far from mandatory. Academics have categorically struggled in our program and entrepreneurs have categorically done really well.

Good question. I think talent vs funding - most people get this fully backwards, with NGOs worrying way more about funding than they should and way less about talent than they should. I also think a lot of charities worry about doing something that causes harm (pretty rare) vs just being really inefficient at accomplishing your goals and effectively wasting resources (very common). NGOs that focus on great talent and getting a lot done per $ end up going a pretty long way.

I am probably more excited about outreach than research when it comes to high absorbancy opportunities like vitamin A vs malaria nets. I am probably most excited about research on things like family planning or livelihoods that we have not as a movement dug as deeply into and I think there are reasonable ethical views that would prioritize the top of those areas over the top of classic direct delivery global health.

I think it starts at an earlier stage than something like CE generally operates. I sometimes think of CEA as like a PhD program - it's best fit for someone who has been working on impact directly for a long time. I think what is missing in most LMICs is the early university and chapter infrastructure to build up networks and knowledge so that there is more ready-to-found/be-Sr-staff talent built up over time. I generally wish more people spent time living in LMICs as I think it does give you a practical edge that is hard to get otherwise.

I don't think size is the best determiner. I think the seed network or mid-stage funding is probably higher EV at the cost of higher risk and more time to assess. I think if a donor finds getting into the weeds and speaking to the charities more fun I think they should go earlier stage. If they are time poor and want a reliable index fund I think it's hard to beat GW/ACE recs.