I think his description of the nature of explanations is relevant here though. Most phenomena, that one might like explained, or wish to understand, aren't as irreducible (at least in terms of things familiar to the questioner) as the fundamental forces!
In terms of education, explaining things at one level of reduction, even if there is more that could be said, meaningfully increases the knowledge of the student (and allows them to ask the next question, should they be interested). This really isn't the same as an analogical "explanation" that in fact explains nothing.
of course, it's easy to say that and hard to come up with good but simple models on the spot, but people don't gain knowledge by learning opaque phrases (that's how we got the whole "quantum woo" epidemic a decade or two ago)
For simpler and more accessible topics though, like the examples I gave, I do think there is value in teaching partially-correct simplifications. For one thing, you have to start somewhere, so if some partial understanding is what sparks some kid's interest into pursuing the sciences, then that is a good thing.
Maybe more importantly(?), given that most people won't be going into science-based jobs, is just having some base level of scientific literacy, especially in a world (particularly the USA) where religious belief and conspiracy theories, rather than rational thought, seem to be in danger of dominating society.
>trying to explain things like quantum physics to the layman (as if scientists themselves really understand it)
There's this common misconception that physicists somehow don't understand quantum mechanics. The problem with quantum mechanics (and a whole host of other experimentally-supported theories) is that they describe things so completely removed from our everyday experience (and indeed, occur in regimes that human beings have never experienced even indirectly until the last 100 or so years) that our everyday intuition is wrong every time.
But if you understand the math of quantum mechanics, and you understand how the experiments relate to the math, then you understand how quantum mechanics works. Its not actually that complicated, its just weird.
There are ~1900 PhDs in physics awarded every year, so certainly, tens of thousands of people across the world do understand how it works. Now, if you're going to quibble and say "do the scientists understand what spin really is?" then congratulations, you've just reinvented Humean skepticism, I guess human knowledge reached its maximum extent in 1748.
OK, but is that really understanding, or just an ability to crank the handle of the model and make a prediction? How should we interpret what is going on? What to make of observational wave function collapse, or long-distance quantum entanglement (which seems completely at odds with our understanding of space and time)?
My view is that, as it pertains to any theory, the ability to make reasonably accurate predictions within a given region, and reasonably accurately describe the region where those predictions are correct is understanding.
I don't believe our knowledge is complete, because the universe is effectively infinite, so there's room for effectively infinite combinations and permutations of phenomenon even within our understanding. But when we encounter something that contradicts our understanding, that doesn't mean we didn't understand anything we experienced before. It just means that our overall understanding was incomplete.