The trouble is you'll never know what the actually numbers were until after the fact - the best you can do is make some educated investments at an 80/20 point. So either you learn yourself or you rely on consensus and authority, etc. You give to a middleman charity that has people who spend the time or do the work based on a theory of what should be done. You're relying on their expertise, and the expertise of those who say that, yes, you should put your money with charity A rather than charity B. But you'll never know if you could have got that 0.2% doing something else. Really, the best you should expect is that you can make a good enough choice to do good versus wasting your donation.

I started supporting SENS pretty early, so I had to actually go read the primary literature and take a few years to figure out that yes, I think that this is the best fulcrum to make progress in changing the research community, building the right tech, etc.

Nowadays, you can actually rely upon more distributed middle man expertise. E.g. that Peter Thiel invested millions in SENS, or that the SRF advisory board includes numerous researchers who are at the tops of their fields, names appearing in the media, all of whom endorse the SENS approach (http://sens.org/about/leadership/research-advisory-board), and so on and so forth.