Of course, that's also not often a choice I (have to) make. I've a set amount of donations to a few regular charities, whereas a situation as above would "count" towards my regular daily expenses rather than charitable givings.
It's probably also related to always being able to do more. Yes, maybe giving €x to one of the Givewell organisations is more effective than giving it to Make-A-Wish, but often it's not a game of either-or. I'm sure a lot of people donate to Make-A-Wish because they encounter it and sympathise with it, rather than spending effort looking for the most efficient way to spend <arbitrary amount designated for charitable donations>, so it's not even a question of effective charity vs. MAW, but MAW vs. sitting there in your savings account.
People only investigate a finite number of things, adding just 1 more to the early stage to the millions already tried is not likely to be useful, but we simply don't know.
PS: After that point much larger sums are involved, but even then only a finite number of tests get done and they are very dependent on funding.
A drug that saves 1,000 lives per year worldwide at the cost of 1 billion dollars seems expensive. But that's 1 million lives over the next 100 years for 10,000$ each and we would still know how to make that drug in 100 years. Dropping that to 1,000$ per life over ~1000 years.
And if you consider 1 billion per 1 thousand lives/year then cancers's 8.2 million deaths/year is clearly worth ~8.2+ Trillion dollars in research to cure or ~4.1 Trillion to cut in half.
PS: Global cancer research only recently hit 100 Billion per year so it's going to take a long time to hit 4.1 trillion yet we already save a large chunk of potential cancer victims.