Future potential (“women and children first”), societal compliance (criminals vs not), etc.
What you think is objectively better with something as open ended as $300m is completely subjective.
One glaring obvious problem with the EA people, is that they double count.
Say I give money for bed nets, and the bed nets save 100 lives.
But say I only gave that money because Eliezer Yudkowsky convinced me to. Was it really him that saved 100 lives? Or do we divide them between us somehow? How many should go to the people who actually distributed the bed nets I paid for?
You'd think the EA people would have good answers to these questions. Or at least, any answers. But they don't. Not any consistent ones. They cheerfully double-count in all their most famous arguments.
At least in your fish example they used it for the intended purpose for some period of time and there is still some low utilization for dried fish. It’s also half of 300 million so it seems they did more with less despite it being a failure.
Regardless, effective altruism is bullshit. They can donate the money however they want. They wouldn’t be a better person or even more charitable if they were an “effective altruist.”