He's right about science education. Way too much of it is about learning facts and calculating numbers from equations. Basically, things that are easily examinable.

I've found the most enlightening science courses have been the ones where you get a story like this:

1) Someone was wondering what shape DNA has, because this might answer questions about how it is copied, how proteins are eventually made from its code, and so on.

2) One way to look at is through x-ray patterns. And this can be as shallow or as deep as you like, there's certainly enough math here if you want to go that way. You can also look at other historical attempts to uncover the structure, ones that failed but nonetheless were considered a reasonable thing to investigate. Plus auxilliary experiments like where it was discovered that phosphoros was part of it.

3) We looked at the x-rays and someone figured out that it is consistent with a double helix, because (evidence).

4) Further questions arising from this double helix hypothesis seem to confirm the structure, and opened up new questions, which we can look at in the same way.

I like the Bill Bryson way of explaining things. He mixes in the people along with the existing ideas, evidence that needed explanation, the breakthrough, and so on.

But my main point is science teaching shouldn't be about the subject so much. You benefit more from understanding how the Krebs cycle was discovered than from memorizing all the parts of it.

For most people by far, the point of learning science it to learn how to apply that style of thinking to all your judgements (should I vaccinate my kids? What do I think of modern monetary theory? How safe is my car?), rather than learning specific things that were discovered in scientific fields.

If you learn how to think the way the article says, you will also be better able to judge how certain you should be about various claims. You'll be asking more what evidence the claims rest on than what are the claims themselves.