While I'm sure there is a lot of thinking behind the scenes not reflected in this NY Times piece, so perhaps what I have to say here is already being considered, one gets the impression that this initiative is looking to make rather large grants (in the tens of millions of dollars range) so the money can be put to strategic use. Such an approach is perfectly understandable from the viewpoint of the donors, who'd like to understand the consequence that results from their gifts as well as to steer outcomes in a more positive direction, when that is possible.
However, just as many commercial startup ventures are starved for capital, so too are many startup not-for-profit charitable endeavors. These organizations have very dedicated and diligent founding members and a cadre of others who volunteer their time in the hope the organization can make a go of it. Fundraising from small donations is the standard way such organizations try to meet their financing needs, which are typically quite modest indeed, and yet often go unmet. Many of these organizations fail. Some might fail even if the funding were adequate - their message doesn't hit the right chord. But many others have a good message for which there is a real audience, yet will fail nonetheless because they can't raise the needed funds.
So my thought is that some small fraction of the Chan Zuckerberg Initiative be devoted to small grants (in the tens of thousands of dollars range) but unlike other Foundation giving of this sort, which targets the areas that are candidates for funding, the Chan Zuckerberg Initiative be unrestricted as to mission of the recipient organizations. Likewise, I would suggest a fairly minimal application process for grant candidates. Allocating the grant funds would not be a search for the proverbial needle in a haystack. It is hard to see how that could make sense, as the costs of sifting through all the applications could end up swamping the benefits from the small grants themselves.
Instead, imagine that the grants were determined purely as a random draw from the applicant pool. Given the high visibility of the Chan Zuckberberg Initiative, one might imagine that pool would be quite large. Even if the success rate on the small grants was quite low, which I assume it would be, the conditions have been set to look at why organizations who receive grants nonetheless fail. (The reporting requirement for grant recipients would be much higher than for initial applicants. One prime reason for the Initiative funding of the grants is to study these recipients.) One might hypothesize that many of these organizations make some basic mistakes that hamper their ability to succeed. If so, and if those mistakes could be identified, future organizations could learn from this and thereby avoid these errors.
The leverage then from such a small grant program, even if it continued with random selection of recipients for funding, is that it might very well influence the behavior of all applicants, not just those that receive the grants. In this way it could raise the success rates of all such startups and in so doing also attract other funders to support these organizations.
Let me close with one other point, which relates this sort of charitable giving to income inequality. It may just be how the piece in the Times was written, but the impression created about the large grants already made is that the recipients were themselves pretty well heeled. In other words, these amount to the uber rich giving to the very well off. Good works will no doubt be done this way. But it blocks many other potential recipients where good works could also be done - these among the poor, working class, and their friends, because they simply aren't networked in a way where they can raise sufficient capital to make them targets for a big grant. These other potential recipients should not be left out in the cold. This suggestion is a way to include them in the process.