An interesting idea, but I think this argument leaves out an important fact: money in the present is more useful than money in the future. This is often captured in accounting by use of a "discount rate". A pretty conventional number is 10% per year: waiting an extra year to receive some money makes it worth 10% less, all else equal. Comparing a typical interest rate and a typical discount rate, I would conclude that you should donate now.
Furthermore, if this is really the best way to do charity... well, charities could just invest the donations they receive and cash out when the time is "right". But since organizations don't die, when would that ever be?
Lastly, the idea that you should invest in a single "best" charity entails the idea that your donation is too small to make much difference in which charity is best. But some problems have already been solved! More problems will surely be solved in the future. It seems that charities could face diminishing returns -- like most other investments -- so your most effective strategy should sometimes be to split among more than one. Especially if you have followed the advice to invest for your entire life, and you do have a huge sum to donate.